UNIVERSITY OF GHANA A Survey of Prescribing Patterns in the Wassa West District of Ghana with Special Reference to Antibiotics Dr William Kofi Bosu BSc, BMedSci, MB ChB A dissertation presented to the School of Public Health in partial fulfilment of the requirements for the award of the Master of Public Health degree Legon January 1996 (j 347499 'RjK\.Z67.‘ Q>(>5 ~jhm n W i TO ROSEMARY, BERRIL AND DARY DECLARATION I hereby declare that this submission is my own work and that, to the best of my knowledge and belief, it contains no material previously published or written by another person nor material which to a substantial extent has been accepted for the award of any other degree or diploma of a university or other institute of higher learning, except where due acknowledgement is made in the text. SUPERVISORS Prof. Ebenezer Laing, Botany Department Mr. Alfred A.D Obuobi, School of Administration ABSTRACT Study title: A survey of prescribing patterns in the Wassa West district of Ghana with special reference to antibiotics. Statement of the problem: The availability, affordability and rational use of drugs constitute a fundamental indicator of quality of health care delivery. Ghana has spent up to 40% of its recurrent budget to ensure the availability of. drugs in the health system. However, irrational prescribing has led to a wastage in drugs supply. 4 - •' < . - - \ & J . In the Wassa West district, several obstacles to rational prescribing exist: These include the lack of an active programme to promote rational prescribing, the lack of objective information to guide prescribers in their practice in the face of an increasing array of new drugs and the aggressive promotional activities of drug companies. A previous rapid qualitative assessment of the pharmaceutical system in the district indicated that there were inappropriate prescribing practices. The absence of relevant, objective and measurable indicators limited the usefulness of the findings of the study as a basis for the evaluation of any interventions instituted to improve the prescribing practices. This study provides the requisite baseline measures of prescribing patterns in the district and identifies areas for targeted interventions. Main objective: To describe and compare the patterns of drug prescribing in government health facilities in the Wassa West district with a view to developing strategies to improve prescribing practices. Place of study: Seven government health facilities located at Tarkwa, Nsuaem, Simpa, Dompim, Bogoso, Huni Valley and Prestea Himan. Study type: The study was cross-sectional comparative in type. Methodology: Retrospective prescribing data covering a one year period from June 1994 to June 1995 were obtained from 700 outpatients' clinical record cards, 100 from each of 7 government health facilities in the district. At the district hospital, prospective data from 100 outpatient prescriptions were also collected because of a relatively large number of incomplete data from the retrospective records. A one-day prevalence survey of prescribing patterns for 45 inpatients at the district hospital was also undertaken. Finally, the type of prescribers and the availability of the national Essential Drugs List (EDL) manual and some common drugs in all 7 health facilities were determined. Findings and conclusions: • irrational prescribing is both serious and widespread in the Wassa West district • prescribing patterns differed significantly between the health facilities. Bogoso and Himan had the worst prescribing indicators • prescribing patterns in the health centres were worse than those in the district hospital perhaps a reflection of the lower cadre of prescribers and a relative lack of diagnostic facilities • treatment records were poorly kept, at the district hospital as a result of a heavy workload • there was an average of 4 prescribers comprising combinations of medical officers, medical assistants, nurses and technical officers in each facility • the number of drugs prescribed per patient ranged from 0-13 with a mean of 4.6 • 80.0% and 57.9% of patients were treated with one or more injectable drugs and antibiotics respectively. The preference for injections was influenced by patient demand as well as by clinical and financial considerations. Antibiotics were commonly prescribed for malaria, respiratory infections and soft tissue infections • 64.6% of all drugs were prescribed by their generic names; 97.1% were on the national Essential Drugs List. The use of essential drags was favoured by procurement from a central Regional Medical Store. Prescription by non-generic names was due to a concern that chemical sellers were not familiar with the generic names and may therefore not be able to dispense the right drugs • there were inappropriate treatment practices. About 87% and 41% of patients with malaria were treated with an injection chloroquine and antibiotics respectively. This was attributed to the lack of refresher training for prescribers in management of some common conditions • the national EDL manual was not available at any of the facilities surveyed owing to a general national shortage of the manuals • except for drugs for skin diseases, drags for the treatment of common diseases were generally available in the facilities • in relation to the severity of illness, drugs for inpatients of the district hospital were greater in number, more expensive and more frequently administered parenterally than those for outpatients. Recommendations: Policy makers: The urgent development, proclamation and promotion of a national drugs policy backed by the relevant legislation should provide a framework for the improvement of prescribing practices and for the entire pharmaceutical sector. The national Essential Drugs Programme should be reactivated and given the necessary legal backing. Funding should be provided for the regular revision and widespread marketing of the EDL manual. Training institutions'. Training institutions should prepare medical officers and medical assistants for the "real-world" practice of prescribing under negative pressures by adopting a problem-based teaching of therapeutics in place of the traditional approach of merely transferring knowledge about the action of drugs in the treatment of diseases. Regional Health Administration'. The regional health authorities should obtain copies of the EDL manual for distribution to the districts, ensure availability of essential drugs in the Regional Medical Store, organise training .as well as assist districts to prepare appropriate treatment guidelines for common diseases. i District Health Administration'. • The district health authorities should establish a District Medical Store to supply drugs to and monitor the consumption patterns of all health facilities. • a District Drugs and Therapeutics Committee should be formed to promote rational prescribing in the district • there should be regular in-service training for prescribers in which the management of common diseases, polypharmacy and the over-use of injections and antibiotics will be discouraged. Funding for training could be obtained from the district's quarterly financial allocation and donor agencies • training should be followed up with a programme of regular supervision of medical assistants and nurses by the medical officers at the district hospital • training of chemical sellers to correctly interpret prescriptions should favour generic prescribing • the district should undertake further studies on the nature and extent of prevailing prescribing influences. Regular assessment of prescribing practices for government as well as mines and private clinics will be necessary. Patients and communities'. Posters on the walls of clinics should be provided to educate public on the effects of patients' demands for certain drugs as well as educate the public on the dangers of self-medication, The appropriate use of drugs should be incorporated into the district's intersectoral social mobilisation programme. ACKNOWLEDGEMENTS I am indebted to the Director and staff of the School of Public Health ("without walls") for the problem-solving oriented training in public health which I received. I hasten to add that the School's provision of notebook computers at subsidised prices to its students has proved invaluable in various academic (and non-academic!) endeavours. Prof Ebenezer Laing and Mr. Alfred A.D Obuobi, my .academic supervisors and Mr. Tom Kumekpor were most helpful in guiding me throughout the study. I was also privileged to benefit immensely from the tutelage of Prof. David Ofori-Adjei of the Centre for Tropical Clinical Pharmacology and Therapeutics. Prof. Richard Biritwum of the Department of Community Health made valuable comments on the early drafts. I am most grateful to the Wassa West District Health Management Team under the leadership of Dr. Teddy Avotri for their support and encouragement for the study. I am also thankful to Dr. Edward Osei Kuffuor of the Tarkwa Government Hospital and all the Medical Assistants in the district for their assistance and co-operation with data collection. I was impressed by the general enthusiasm expressed by prescribers in improving their practices. The UST School of Mines kindly provided me with accommodation at a subsidised rate throughout the period of my stay in Tarkwa. I acknowledge expert software assistance from Mr. Tony Williams, formerly of the Health Research Unit in the design of my data collection instruments. I thank all other health workers who contributed in various ways to making the study worthwhile. Very many thanks to the Regional Health Admin., Cape Coast and to my sponsors, the Mnistry of Health who made my training in public health possible and enjoyable. Finally, I am profoundly grateful to my wife, Rosemary and my family for their love, patience and encouragement and all at each stage of my scholastic pursuit. CONTENTS EXECUTIVE SUMMARY ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES LIST OF ABBREVIATIONS 1.0 INTRODUCTION 1 1.1 Background information 1 1.2 Statement of the problem 2 1.3 Description of study area 4 1.4 Study assumptions 7 1.5 Objectives 7 1.5.1 General objective 7 1.5.2 Specific objective 7 2.0 LITERATURE REVIEW 9 2.1 Economics of drug consumption 10 2.2 Problems in drug supply 10 2.3 Rational prescribing of drugs 12 2.3.1 Polypharmacy 13 2.3.2 Generic name prescribing 14 2.3.3 Prescribing of injections 15 2.3.4 Misuse of antibiotics 16 2.3.5 Inappropriate treatment 20 2.4 Factors influencing drug prescribing 21 2.5 Strategies to improve prescribing behaviour 24 2.5.1 Essential Drugs Programme Activities in Ghana 26 2.6 Methodological review 30 2.7 Conclusion 32 3.0 METHODOLOGY 33 3.1 Study type 33 3.2 Study period 33 3.3 Operational definition of terms 33 3.4 Definition and measurement of variables studied 34 3.5 Study population 34 3.6 Sampling 3 6 3.6.1 Sampling method 36 3.6.2 Sample size 37 3.7 Data collection 40 3.7.1 Permission to proceed 40 3.7.2 Data collection tools and techniques 40 3.7.3 Pretest 41 3.8 Data processing and analysis 42 3.9 Advantages and limitations of study methodology 42 3.10 Ethical considerations 44 4.0 RESULTS 45 4.1 Demographic data 45 4.2 Diseases encountered 45 4.3 Prescribing indicators 47 4.3.1 Number of drugs prescribed per encounter 47 4.3.2 Prescription of drugs on the Essential Drugs List 50 4.3.3 Generic name prescribing 50 4.3.4 Prescription of injectable drugs 51 4.3.5 Antibiotic prescribing 52 4.3.6 Appropriateness of treatment 56 4.4 Case reports 58 4.5 Prescribing indicators from prospective data 60 4.6 Type of prescribers 63 4.7 Health facility indicators 64 4.8 Summary of results 66 5.0 DISCUSSION AND CONCLUSIONS 68 5.1 Number of drugs prescribed per patient 68 5.2 Drugs prescribed from the Essential Drugs List 71 5.3 Generic name prescribing 72 5.4 Prescription of injectable drugs 73 5.5 Antibiotic prescribing 75 5.6 Appropriateness of treatment 78 5.7 Type of prescribers 78 5.8 Availability of Essential Drugs List manual 79 5.9 Availability of drugs for treating common conditions 79 5.10 Summary of discussion 80 6.0 RECOMMENDATIONS 82 REFERENCES 88 ANNEXES 98 Annex 1 Sample size determination for different types of drug use studies 98 Annex 2 Questionnaire for prescribing patterns in health facilities of the Wassa West district with special reference to antibiotics 99 Annex 3 Questionnaire for selected drug-related health facility indicators 101 Annex 4 Drug use at OPDs of District Hospital and Health Centres in the Wassa West district 102 Annex 5 Drug prescriptions for 45 inpatients of the Tarkwa Government Hospital 104 Annex 6 Inappropriate treatment - malaria and malaria with helminthiasis 105 Annex 7 Inappropriate treatment - fever and urinary tract infection 106 Annex 8 Indications for antibiotic use in health facilities of the Wassa West district 107 Annex 9 Price list of selected drugs in the Wassa West district - October 1995 108 LIST OF TABLES Table 1.1 Characteristics of government health facilities in the Wassa West district 6 Table 1.2 Ranks of top 5 diseases from MOH institutions in the WWD in 1993 6 Table 3.1 Definition and measurement of selected variables 3 5 Table 3.2 Sample size calculation for a single health facility 3 8 Table 3.3 Data collection tools and techniques 40 Table 4.1 Age and sex distribution of patients attending government health facilities 45 Table 4.2 Diseases encountered in 700 OPD clinical records 46 Table 4.3 Distribution of prescribing indicators 47 Table 4.4 Number of drugs per prescribing encounter 48 Table 4.5 Encounters in which no drugs were prescribed at the Tarkwa Government Hospital 49 Table 4.6 Frequency of generic prescribing of common drugs 51 Table 4.7 Top 5 injectable drugs prescribed 52 Table 4.8 Indications for antibiotic use in the district hospital and health centres 55 Table 4.9 Cost of antibiotic use in relation to other drugs 56 Table 4.10 Prescribing indicators from prospective data at the District Hospital 60 Table 4.11 Distribution of type of prescribers in government health facilities 64 Table 4.12 Availability of some common drugs in health facilities 65 LIST OF FIGURES Fig. 1 Map of the Wassa West district showing sub-districts 4a Fig. 2 Factors influencing prescribing behaviour 21a Fig. 3 Commonly prescribed drugs at OPDs of health facilities 48 Fig. 4 Cases with injectables or antibiotics in government health facilities in the Wassa West district 53 Fig. 5 Antibiotic prescribing - District Hospital 53 Fig. 6 Antibiotic prescribing - Health Centres 54 Fig. 7 Antibiotic prescribing - Prospective OPD patients at the District Hospital 62 Fig. 8 Antibiotic prescribing - Inpatients at the District Hospital 63 L IS T O F A B B R E V IA T IO N S AT£ A ntitetanus serum BHC B ogoso H ealth Centre DANIDA Danish International Developm ent Agency DHC Dompim H ealth Centre DHM T District H ealth M anagement Team EDL Essential D rugs List EDLIZ Essential D rugs List o f Zimbabwe EDL/NF Essential D rugs List and National Formulary EDP Essential D rugs Programm e HHC Himan H ealth Centre HVHC Huni Valley Health Centre IM i Intram uscular [NRUD International N etw ork for Rational use o f D rugs MA M edical Assistant MOH M inistry o f Health M OH & CW M inistry o f Health and Child W elfare MSH M anagem ent Sciences for Health NHC Nsuaem Health Centre NRCD National Redem ption Council D ecree OPD Outpatients' Department RMS Regional Medical Store RPM Rational Pharmaceutical M anagem ent SHC ( Simpa Health Centre TGH Tarkw a Governm ent Hospital UK United Kingdom URI U pper Respiratory Infections USA United States o f America USAID United States Agency for International Developm ent UST University o f Science and Technology UTI Urinary Tract Infection WHO W orld Health Organisation W \yD W assa W est District ZEDAP Zimbabwe Essential D rugs Action Program m e CHAPTER ONE 1.0 INTRODUCTION 1.1 Background information Drug management forms a vital component of any health system. While drugs alone are not sufficient to provide adequate health care, they do play an important role in the diagnosis, prevention, alleviation and cure of diseases as well as for the maintenance of health. The provision of essential drugs is one of the pillars of the Primary Health Care concept which was adopted by Ghana in 1978. Public satisfaction with health services depends on among other factors, the maintenance of a regular, affordable and accessible supply of appropriate drugs (Waddington & Enyimayew, 1989). Poorly managed drug systems also affect the morale and satisfaction of service providers who may be frustrated by chronic shortages of needed drugs. The analysis of drug supplies and consumption patterns therefore, can be used to monitor the quality of health care (WHO, 1994). Until 1992, about 30-40% of the total recurrent budget for health in Ghana was spent on drug procurement (Fofie, 1994). In spite of the immense value of drugs to the public health system, many developing countries have problems ensuring a regular supply of drugs. Further, even when drugs are available, a large proportion of it is wasted through irrational prescribing, pilferage -and inefficient management (Foster, 1991). In its assessment of the pharmaceutical secfor of Ghana, the National Drugs Committee (1994) stated that "the real blow to health finances comes at the end of the line, where overprescribing reaches a level almost unequalled in Africa and wastes a large proportion of scarce funds". Recent efforts in Ghana to rationalise drug management and to ensure the availability of drugs have included the introduction of an Essential Drugs List (EDL) in 1988 and i Introduction/1 the Cash and Carry fell-cost recovery system in 1992. These have still met with various managerial problems and have done little to address the problem of irrational drug prescribing. The district level is the most appropriate level for implementing primary health care (WHO, 1994). Accordingly, Ghana has directed its efforts at improving the functioning of the district health system with the implementation of the Primary Health Care strategy and the Strengthening District Health Systems initiative (MOH, 1992). In furtherance of its decentralisation policy, the central government now allocates fends for health development directly to the districts. The introduction of the Cash and Carry system has enabled districts to exercise a greater control over their drugs supply problems. With these elements, district health systems are increasingly directing their efforts at improving their drug management systems and the quality of health care delivery to their populations. It is suggested that the investigation of drug prescribing practices at the district level is one important strategy that will contribute to meeting these objectives. 1.2 Statement of the problem Irrational prescribing remains a problem in many developing countries (Isenalumhe & Oviawe, 1988; Mnyika & Killewo, 1991). A recent Rational Pharmaceutical Management (RPM) Project assessment of the pharmaceutical sector of Ghana identified irrational prescribing as a major problem in the public health system. The study covered 20 health facilities in 5 regions, namely Brong-Ahafo, Greater Accra, Northern, Volta and Western regions. The health facilities in each region comprised the Regional Hospital, a District Hospital and within the district, two health centres (Rankin et al., 1993). These facilities were selected out of 10 Regional Hospitals, 36 District Hospitals and 252 Health Centres/Posts in the country (MOH Annual Report, 2 Introduction/1 1993) using a WHO/INRUD methodology for investigating drug use in health facilities (WHO, 1993). The study was intended to provide a national picture of drug use and so did not assess prescribing practices in the different types of facilities. Different facilities have varying influences such as drug availability, type of prescribers and available diagnostic facilities on prescribing behaviour . Therefore, they are best studied separately for the purpose of designing appropriate remedial interventions to improve prescribing. The drug prescribing practices of a district health system have not previously been studied. In the Wassa West district, prescribers in peripheral health facilities often are isolated and lack objective information on rational prescribing. Their main sources of information are drug representatives and commercially-oriented publications. The Cash and Carry system is a potential obstacle to rational prescribing. The system permits health institutions to buy their own drugs according to their own perceived needs which are likely to be dictated by existing prescribing habits. Drugs including those that are not on the EDL may be purchased from the private sector. There is also the problem of the prescriber having to select from an increasingly large number of drugs on the market. The district has no defined programme to improve the prescribing practices of clinicians. Further, the availability of the Essential Drug List and National Formulary in the district health facilities has not been assessed. A 3-member ad hoc District Formulary Committee comprising the District Pharmacist, a Senior Nursing Officer and a Medical Assistant was formed in 1994 to rapidly assess the drug management system in the district with regards to procurement procedures, storage and use of drugs in government health facilities. Prescribing practices were not systematically studied. Prescriptions were analysed only to determine whether prescribed drugs were appropriate for stated diagnoses. 3 Introduction/1 The committee's major findings were that stock levels of essential drugs at the health institutions were inadequate; there was significant wastage due to inappropriate prescribing and drug purchases; there was irregular supply of drugs and drug requisitioning was not related to disease patterns (Report of Wassa West District Formulary Committee, 1994). Unfortunately, these findings were not quantified and therefore provide insufficient basis on which to evaluate any remedial measures to improve the situation. It is expected that the present study will provide baseline measures of prescribing patterns and identify areas for targeted interventions. These measures will enable any interventions to be evaluated. They may also provide baseline information upon which the impact of drug use may be assessed. For example, antimicrobial susceptibility patterns may be related to the extent and pattern of antibiotic use. Finally, the findings may provide useful information for the development of treatment guidelines. 1.3 Description of study area The Wassa West district is one of eleven districts in the Western Region of Ghana. It is located in the central south of the region and covers an area of about 2,450 square kilometres. It has a population of about 208,000 and a growth rate of about 3.1%. The district is made up of 209 communities of which about 88% is rural. It is administered by the Wassa West District Assembly which is headed by a District Chief Executive. Tarkwa is the administrative capital. The principal occupations are farming, mining and trading. Health services in the district are managed by the Wassa West District Health Management Team. All manner of health services are available in the district. They include services provided by traditional and spiritual healers as well the private, missionary and governmental practitioners. The health district is divided into 6 sub- 4 KEY 1 PRESTEA I BOGOSO 3 INSU,'/ HUN1 VALL̂ Ii 4 TARKWA 5 DOMPIM 6 8ENS0 I NSUAEM m s SAW WEST SUB DISTRICTS ________________Fig. 1 4 a Introduction/1 districts, namely Nsuaem, Dompim, Tarkwa, Bogoso, Huni Valley and Prestea Himan (fig. 1). The District Hospital is located in the Tarkwa sub-district. The other sub­ districts are subserved by Health Centres corresponding to their names. Additionally, the Dompim sub-district has one other health centre at Simpa. The major health institutions in the district include the following: • 1 District Hospital at Tarkwa • 3 Mines Hospitals at Tarkwa, Nsuta and Prestea • 6 Health Centres at Nsuaem, Simpa, Dompim, Bogoso, Huni Valley and Himan • 3 Mines Clinics at Tarkwa (2) and Bogoso • 2 Industrial Company Clinics at Bonsa (rubber) and Aboso (glass) • 6 Community Clinics at Aboso, Awudua, Amantin, Huni Ano, Nyarso and Iduapriem • 4 Private Clinics The district hospital and health centres are the main government health institutions in the district and are therefore under the direct control of the DHMT. Apart from two quasi-government Mines Hospitals at Nsuta and Prestea, all the remaining health institutions are privately-owned and are therefore responsible for selecting their own staff and procuring their own supplies. The two quasi-government hospitals are managed independently of the DHMT. The Mines and Industrial Clinics generally cater for the company staff and in some cases their dependants. The health centres are in the charge of medical assistants as is one community clinic at Aboso. The medical officers at the district hospital and the medical assistants at the health centres are the major prescribers in the government health facilities. The medical assistants at the health centres are frequently assisted by some of their nursing staff. The district also has two pharmacists who both work at the district hospital. Most of their drag supplies for the health facilities are obtained from the Regional 5 Introduction/1 Medical Store at Sekondi. The DHMT is currently in the process of establishing a District Medical Store to cater for the district hospital and health centres. There are also 43 Licensed Chemical Shops in the district from which patients may purchase drugs. Some characteristics of the government health facilities for 1994 are presented in table 1.1. Only the district hospital has facilities for admissions as well as for laboratory and radiographic diagnosis. Additionally, the district hospital has 4 main wards namely the Female, Children's, Male Medical and Male Surgical Wards. Other units of the hospital include a Dental Clinic and two theatres. Table 1.1 Characteristics of government health facilities in the Wassa West District Characteristic Tarkwa Gov. Hosp. Nsuaem H/C Simpa H/C Dompim H/C Bogoso H/C Huni Valley H/C Himan H/C Location Tarkwa Nsuaem Simpa Dompim Bogoso Huni Valley Himan Year Commissioned >100yrs ago 1990 1991 1990 1987 1990 1984 Catchment Pop. 81755 11401 21210 21210 21210 21210 31344 Staff Strength 103 30 8 3 10 11 18 Staff:Pop. Ratio 1:794 1:380 1:2651 1:7070 1:2121 1:1928 1:1741 No. of beds 69 18 10 12 19 9 6 No.of Outreach Pts 13 10 nil 3 7 8 13 OPD Attdce 1994 23500 3074 3185 2055 7018 2674 8948 The top 5 diseases in the government health facilities in 1993 are shown in Table 1.2. Malaria accounts for about 45% of all cases seen. Table 1.2 Ranks of top 5 diseases from MOH institutions in WWD in 1993 Disease Tarkwa Gov.Hosp. Nsuaem H/C Simpa H/C Dompim H/C Bogoso H/C Huni Valley H/C Himan H/C Malaria 1 1 1 1 1 1 1 Diarrhoea Dis. 2 4 2 4 5 3 4 Upper Respiratory Infections 3 2 4 2 4 5 2 Diseases of Skin 4 5 5 5 4 5 Eye Infections 5 - - Accidents 3 3 3 3 2 Pregnancy-related Disorders 2 Intestinal Worms - - - - 3 6 Introduction/1 1.4 Study assumptions The following assumptions have been made to guide the study: • there is irrational drug prescribing in the Wassa West District • irrational prescribing is more prevalent in health centres than in the district hospital • many drugs are prescribed for each patient • there is over-use of antibiotics and injections • drugs are frequently prescribed by their brand names rather than by their generic names • drugs not on the national Essential Drugs List are frequently prescribed • inappropriate treatment prescriptions are common • prescribers have little access to the Essential Drugs List and National Formulary to guide their treatment decisions 1.5 Objectives 1.5.1 General Objective The general objective of the study was to describe and compare patterns of drug prescribing in government health facilities in the Wassa West district with a view to developing appropriate strategies to improve prescribing practices and ultimately, the quality of health care delivery. 1.5.2 Specific objectives The specific objectives of the study were as follows: 1. To compare the outpatient drug prescribing patterns between different government health facilities in the Wassa West district 2. To describe the prescribing patterns for inpatients of the district hospital 7 Introduction/1 3. To estimate the following measures: • average number of drugs prescribed per patient • the proportion of patients receiving antibiotics • the frequency of specific antibiotics prescribed • the cost of prescribed antibiotics in relation to the total cost of all drugs • the proportion of patients receiving injectable drugs • the frequency of specific injectable drugs prescribed • the number of injection episodes prescribed per patient • the proportion of drugs prescribed which are on the Essential Drugs List • the proportion of drugs prescribed by their generic names • the frequency of specific drugs prescribed as generics 4. To describe the route of administration of prescribed drugs 5. To list the indications for antibiotic use in the district 6. To assess the appropriateness of prescriptions 7. To describe the type of prescribers working in each health facility 8. To assess the availability of the Essential Drugs List and National Formulary at the health facilities 9. To investigate the availability of drugs used to treat common conditions at each health facility 10. To make recommendations to improve prescribing practices 8 CHAPTER TWO 2.0 LITERATURE REVIEW The management of pharmaceuticals, particularly to ensure the availability of drugs is an essential component of any health system. This is because of the vital role drugs play in the management of health problems. Public confidence in health workers and satisfaction with health services depend upon the maintenance of a reliable and affordable supply of drugs. The availability, accessibility, affordability and appropriate use of drugs therefore constitute a fundamental indicator of the quality of care in health facilities (WHO, 1994). In a qualitative study among various community groups in the Ashanti-Akim district of Ghana on the utilisation of health services, Waddington and Enyimayew (1989) found that quality of health care was a more important determinant than its cost. The quality i of care was seen in terms of the perceived quality of medical care, staff attitudes and the availability of drugs. Many patients expressed their resentment at being charged a consultation fee and receiving a prescription at the government health facility but no drugs. Unfortunately, even when drugs are available, many of them are wasted because of inappropriate prescriptions and use. The judicious management of drugs is particularly important in Africa and the developing world where various human experiences are medicalized i.e. defined as medical problems (Haaijer-Ruskamp & Hemminki, 1993). For instance, ageing is viewed as a medical problem for which analgesics and multivitamins are believed to be essential. In Ghana, indomethacin is popularly known as aberewa bebo ball meaning i the "old lady will play football" reflecting the belief that taking the drug will make the elderly strong enough to play the game of football. Drug use has increased as a 9 Literature review/2 consequence of medicalization whilst at the same time the production of effective drugs has itself contributed to the process of medicalization (Bell, 1987). 2.1 Economics of drug consumption At both private and public health facilities in Africa, pharmaceutical expenditures typically make up 20% to 30% of total recurrent costs, ranking second only to personnel costs (World Bank, 1994). The World Health Organisation (WHO) reckons a much higher expenditure by developing countries; it estimates that up to 40% of the total health care budget may be spent on drugs (WHO, 1977). In Ghana, a sum of {532.4 million (cedis) representing about a third of the total recurrent budget for 1976-1977 was allocated to supplies. The corresponding allocation for wages and salaries was 043.5 million. The proportion of recurrent budget allocated to drugs for the same period was estimated to be about 24% (Barnett el al., 1980). Until 1992, drugs were estimated to account for about 40% of the total health budget of Ghana with more than 80% of this expenditure being in foreign currency. In addition, drugs accounted for more than 60% of the recurrent expenditure of the health budget apart from wages and salaries (MOH, 1989). Since the introduction of the Cash and Carry system in 1992 however, there has been no public sector budget for pharmaceuticals (Fofie, 1994; National Drugs Committee, 1994). In 1993, MOH was able to finance 70% of its drug requirements from Cash and Carry operations (MOH Annual Report, 1993). 2.2 Problems in drug supply The major problems facing pharmaceutical systems in developing countries include limited economic resources, lack of organised drug policies, shortage of trained health personnel and inefficiency and waste in drug supply. Patients attending public sector 10 Literature review/2 facilities in Sub-Saharan Africa may be receiving the benefits of only $12 worth of quality drugs for each $100 spent because of widespread inefficiencies in drug management (Foster, 1991; WHO, 1977; World Bank, 1994). The factors that contribute to waste in drug supply in Sub-Saharan Africa include the following: • inadequate buying policies and practices - a WHO study of buying practices in Nigeria found that by shifting from brand name to generic drugs, costs could be reduced by another 25% (WHO, 1988a) • poor quantification of drug needs - estimate of needs are not related to morbidity data and appropriate treatment schedules. In Ghana, when quantification of drug needs of the public sector based on morbidity was carried out by the Ministry of Health (MOH), the overall estimated needs came to approximately $4 million - only half of what actually had been spent on drugs (cited by Foster, 1991) • excessive purchase of low priority drugs ties up the already meagre funds with a potential risk of their expiry unused. In Ghana, a nation-wide audit of drug stocks in 1988 showed that 45% of the drugs were either expired (33%) or due to expire within the next 6 months (12%) (MOH, 1988) • inefficient procurement - procurement is rarely based on competitive bidding for generic drugs. Bulk purchasing of carefully selected essential drugs was estimated to be able to lead to a saving of nearly 40% of the annual drug bill for church health institutions in Kenya in 1985 (Hogerzeil & Moore, 1987) • inefficient storage and distribution - drugs are wasted due to poor storage conditions, poor needs estimation or inventory control and pilferage by employees or theft by outsiders • irrational prescriptions have been estimated to lead to about 50% waste in drug supply to patients. Irrational prescribing practices are further discussed below • non-compliance by patients - the incorrect use of drugs by patients reduces the proportion of drugs that are used effectively by 20% (World Bank, 1994) 11 Literature review/2 In response to the above problems, many countries have implemented Essential Drugs Programmes (EDPs) in order to improve the management, availability and the rational use of drugs. 2.3 Rational prescribing of drugs Rational prescribing requires the basic tools of right drugs, right information and advice. It must be set in the wider context of the rational use of drugs which extends beyond rational prescribing practices and patients' attitudes to include policy issues such as selection, legislation and procurement, all of which influence the availability, source, quality and use of drugs (van der Geest, 1987; WHO, 1992b). The Director-General of WHO has said that "the rational use of medicinal drugs is critical to the most important primary and preventive health care measures; ... unless, there is a regular supply of safe and effective drugs, public trust and interest in primary health care will rapidly deteriorate" (Nakajima, 1992). This statement clearly emphasises the importance of the rational use of drugs in national health systems. i Interestingly, the pattern of drug prescription is similar in developing countries in spite of varying cultures, health systems and economies. There are reports of polypharmacy, prescription of brand-name drugs, frequent use of injections and antibiotics and inappropriate treatment practices in Ghana, India, Nigeria, Tanzania, Kenya and Ethiopia (Barnett et al., 1980; Greenhalgh, 1987; Isenalumhe & Oviawe, 1988; Mnyika & Killewo, 1991; Nabiswa & Godfrey, 1994; Sekhar et al., 1981). These examples of irrational prescription of drugs and their consequences are discussed separately below. i 12 Literature review/2 2.3.1 Polypharmacy The problem of polypharmacy in developing countries betrays how little progress, if any, has been made in the rational use of drugs. A study of outpatient consultation data at the Achiase Health Centre in the Birim District of Ghana's Eastern Region in 1976 indicated an average of 3.9 items per prescription (Barnett et al., 1980). More recently, the USAID-supported Management Sciences for Health (MSH) study on Rational Pharmaceutical Management (RPM) in Ghana reported an average of 4.3 drugs per curative encounter in 20 health facilities in 5 regions in Ghana (Rankin et al., 1993). In Nigeria, the average number of drug items prescribed for 145© paediatric '% ' 'V- 4 outpatients at the University of Benin Teaching Hospital in 1985 was 4.7. Over 76% of the prescription orders contained 4 or more drug items. The average number of drug items for 299 children with diarrhoea was 5.6 (Isenalumhe & Oviawe, 1988). The worst example of overprescription has been reported from Mali where the average prescription contained 10 drugs. These drugs sometimes included a duplication of an ^antibiotic under different brand names. No other details of the survey such as the stu(ly locatioh of type of facility in which the prescriptions were made were provided (cited by Foster, 1991). Another report indicated that 44% of outpatient prescriptions at the national teaching hospital in Nairobi contained 3 or more drugs (Maitai & Watkins, 1980). Many years later, a study covering 4 urban health facilities in Eldoret in the north-western highlands reported the mean number of drugs per encounter as 2.02 (Nabiswa & Godfrey, 1994). Clearly, any comparisons of prescribing data should take account of the different types of health facilities (which determines the type of prescribers, 13 Literature review/2 availability of diagnostic facilities etc.), the number of facilities covered and whether the study unit includes reattendants or new patients only. The higher the number of drugs prescribed per patient, the higher the costs of drug therapy and the greater the chance for adverse drug reaction or interaction. Isenalumhe & Oviawe (1988) observed that the greater the number of drugs per patient, the greater the likelihood of his failure to procure them as a result of economic implications of such prescriptions. While 85% of patients for whom 2 or 3 drugs had been prescribed collected their drugs, only 55% of those with 4-5 drugs and 31% of those with 6-7 drugs collected theirs. Isenalumhe & Oviawe (1988) also observed that many physicians who prescribed multiple drugs hardly allowed patients to relate their complaints before they wrote out prescriptions. Such physicians did not take time to explain anything, including the treatment process to their patients. Their hasty practice had obvious implications for patient compliance. Polypharmacy could lead also to poor compliance because patients get confused about having to take many drags at different times (Boyd et al., 1974). Finally, the use of several unnecessary drugs could deprive patients who genuinely need the drugs from getting them. 2.3.2 Generic name prescribing Prescribing by generic names ensures that the lowest cost generic product available can be dispensed. Accordingly, it is mandatory for all government and government- aided institutions in Zimbabwe to prescribe by generic name only (WHO, 1988c). Brand name drugs are about 5 times more expensive than their generic equivalents (Foster, 1991), In a large health centre in Mali, the price paid for a brand name ampicillin was 2.5-3 times the international generic price. A change in the 14 Literature review/2 procurement strategy for this one drug alone would have led to a 15-20% in the drugs budget (Foster, 1991). Similarly, Mariko (1991) reported that the price of drugs at the Bankass and Koro health centres where drugs are purchased as generics from a non-profit European wholesaler was a third to a quarter of that of comparable drugs at the State-owned People's Pharmacy of Mali which procures its supplies (mostly specialities) by direct contract from local and external markets. A change in procurement policy would have led to a 40% saving on drug expenditures. In the RPM study in Ghana, 59.4% of drugs were prescribed by generic names. Similar studies conducted by the International Network for Rational use of Drugs (ENRUD) in ten countries, namely Yemen, Uganda, Sudan, Malawi, Indonesia, Bangladesh, Zimbabwe, Tanzania, Nigeria and Nepal showed an average of 66.7% generic prescribing (Rankin et al., 1993). In Tanzania which has an Essential Drugs Programme (EDP), the proportion o f drugs prescribed by their generic names in 20 dispensaries in the Dar es Salaam region was 83.9% (Massele et al., 1993). Interestingly, within the Mbeya region of the same country, Mnyika and Killewo (1991) found that almost all health workers prescribed drugs by their brand names although the drug list in their EDP kit used generic names. Similarly, Greenhalgh (1987) also found in a study of the prescriptions of private general practitioners in India that about 96% of all drugs were prescribed by brand names. 2.3.3 Prescribing of injections The frequent, often unwarranted use of injections has been described in several reports from developing countries. In the Achiase Health Centre study in Ghana described Literature review/2 above, 96% of 305 visits were treated with at least one injection, and the average number of injections per consultation was 0.99. The commonest injections were an antimalarial and a penicillin (Barnett et al., 1980). The over-use of injections appears to be a particular problem for Ghana. In the RPM study, 55.7% of patients received injections compared with an average of 24.7% (range 0.2-48%) for the 10 INRUD study countries (Rankin et al., 1993; WHO, 1993). In Tanzania, only 24.6% of cases received injections (Massele et a l, 1993). In Yemen, the proportion of cases with injections in an EDP area and a non-EDP area was 24.8% and 57.8% respectively (Hogerzeil et a l, 1989). It has to be noted that apart from the frequent use of injectable drugs, the number of injection episodes for each injectable drug is also believed to be high judging from the prescribing patterns of such common drugs as chloroquine and the penicillins (Marfo et a l, 1995; Ofori-Adjei, 1989). Unfortunately, there are hardly any reports on the frequency of injection episodes. The over-use of injections not only leads to higher cost for the patient, but also makes extra demands on health institution in terms of trained staff, staff time, sterilisation equipment etc. The potential risk of adverse reaction is much higher with injections than with other routes of administration. Other problems associated with injections include injection abscesses, thrombophlebitis, poliomyelitis, hepatitis B and C and AIDS. 2.3.4 Misuse of antibiotics In Ghana, as in other developing countries, infectious diseases of bacterial origin constitute a major cause of morbidity and mortality. Whereas many of them can be prevented through improved nutrition, immunisation, sanitation and water supply, 16 Literature review/2 many still have to be combated with antibacterial drugs. The use of antibiotics must be based on sound scientific policies for a successful and sustainable fight against infectious diseases. Unfortunately, this is often not the case; the rampant misuse of these drugs has been well documented. The proportion of hospital inpatients who receive antibiotic therapy has been variously reported as 21-28% in Britain (Cooke et al., 1983; Moss et al., 1981); 25-36% in Australia (Raymond et a l, 1989); 28% in South Africa (Till et al., 1991) and 57% in Bangladesh (Hossain et al., 1982). For outpatients, the proportion has been 38.6% in Tanzania (Massele et al., 1993); 46.6% in Ghana; 43.2% for 10 INRUD study countries (Rankin et al., 1993) and 54.3% in Zimbabwe (Stein et al., 1984). The antibiotics were reported to have been used inappropriately in 12.3% of therapeutic prescriptions in Zimbabwe, 2.5-19.0% in Britain, 38.8% in South Africa, 24-39% in Australia and 50% in Bangladesh depending on authors' own criteria for assessment. The proportion of inappropriate prophylactic prescriptions was generally higher than that of therapeutic prescriptions. In an earlier review, Buckwold and Ronald (1979) reported that antibiotics had been used inappropriately or irrationally in 38-66% of cases. Up to 80% of irrational use was given as prophylaxis particularly for surgical cases. Roberts and Visconti (1972) found that 77% of their hospital's antibiotic cost was spent for irrational therapy. In addition, this therapy was responsible for 92% of all patients who experienced adverse drug effects. Other examples of irrational prescribing have been personally observed by the author in his clinical practice in Ghana. These include the use of antibiotics in upper respiratory infections (URI), asthma, malaria, abrasions and clean operations (e.g. breast and groin hernia operations). Unfortunately, there are only few published reports that document such misuse. 17 Literature review/2 In one survey in USA, nearly 60% of physicians used antibiotics to treat common cold (Stolley et al., 1972). Williams (1986) observed that the community practitioner in UK used antibiotics to treat viral upper respiratory infections and terminal phases of malignant disease. At first level health facilities in the Central Region of Ghana, 80.6% of health providers prescribed antibiotics to children with acute URI (Afari et al., 1995). Marfo et al., (1995) in a study on the management of malaria in health facilities in the Cape Coast municipality found that prescribers frequently added drugs such as antibiotics to their antimalarial therapy. In developing countries, chloramphenicol has been widely used for the treatment of diarrhoea, which is often attributed to typhoid fever without bacteriological evidence (WHO, 1983). Loefler (1989) observed that in his travels through Africa, he often encountered hospitals where everyone with a wound was given courses of antibiotics and yet infection rates there were still about 30% or higher. He dismissed such arguments as perceived high infection rates due to lack of sophisticated facilities in theatres, presence of flies and unhygienic habits of patients put forward to justify the misuse of antibiotics. Other examples of 'bad practice' have included the use of topical gentamicin and the use of fusidic acid alone (Davey, 1993). The considerable resources spent on antibiotics further justify the need for appropriate use of these drugs. Antibiotics often account for 15-30% of drag expenditures, the largest share of any therapeutic group of drags (Col & O'Connor, 1987; WHO, 1988c). In developing countries, the lack of adequate resources and the lack of restrictions on antibiotic use coupled with self-prescribing of over-the-counter drugs has led to a situation of sub-optimal use and poor utilisation. On the other hand, in the developed countries, the major issues in antibiotic utilisation have involved the selection of the most efficacious drug and the dose and duration of therapy (Levy et a l, 1987). 18 Literature review/2 The over-use of antibiotics has led to the development of resistance, increase in nosocomial infections and increase in the use of expensive antibiotics. Apart from the increased risk of side-effects, the use of antibiotics could also lead to a reduction in the amount or balance of normal flora in the bowel and other sites. The consequences could be the development of gastrointestinal disturbance, supercolonisation with resistant bacteria or yeast and perhaps in reduced host resistance to infection as normal symbiotic protective flora are removed. The selection of resistant strains is favoured by the widespread use of an antibiotic. This resistance may be lost when the antibiotic is discontinued. In one dramatic report, a severe outbreak of Klebsiella aerogenes infection in a regional neurosurgical intensive care unit in Scotland resulted in 8 deaths from meningitis and 3 from pneumonia. Isolation of infected cases and the use of massive doses of colistin failed to control the outbreak. Only when all antibiotics were stopped both therapeutically and prophylactically was the outbreak brought under control (Price & Sleigh, 1970). Similarly, carbenicillin-resistant strains of Pseudomonas aeruginosa in a bums unit were eradicated by stopping the prescribing of carbenicillin and restricting the use of other antibiotics (Lowbury et al., 1972). Recent reports in Ghana indicate disturbing antimicrobial susceptibility patterns of urinary pathogens and Neisseria gonorrhoeae. More than 75 % of urinary isolates were resistant to the commonly prescribed ampicillin, tetracycline and cotrimoxazole. In contrast, more than 80% of all isolates were sensitive to the more expensive cefuroxime (Newman, 1990; Adjei, 1993). Similarly, for N. gonorrhoeae, up to 100% were resistant to commonly prescribed drugs such as penicillin, cotrimoxazole and tetracycline. Here too, over 90% of isolates were sensitive to norfloxacin, cefuroxime and ceftriaxone (Addy, 1994; Adu-Sarkodie, 1993/94). Similar results have been reported from Nigeria (Ako-Nai et al., 1993). The authors blamed the observed high levels of resistance on the lack of restriction on availability and use of antibiotics 19 Literature review/2 especially in relation to self-prescribing and over-the-counter sales as well as the lack of appropriate standardised treatment regimens. 2,3.5 Inappropriate treatment While the widespread practice of polypharmacy, paucity of generic prescribing and over-use of injections and antibiotics in themselves are examples of inappropriate treatment, in this section more specific clinical examples which bring together these practices are discussed. In broad terms, inappropriate treatment includes the prescription of a drug when not indicated, selection of an inappropriate drug, over- or under-prescribing, incorrect duration of treatment, route of administration or dosing frequency and polypharmacy. In the Achiase Health Centre study in Ghana, antimalarials were used even for cases other than fever or malaria (Barnett et al., 1980). Similarly, in Kenya, antimalarials were used in 396 patients in the non-endemic region of Eldoret without any record of fever or positive blood smear. Twenty-four of 39 (61.5%) children with short attacks of diarrhoea received antibiotics (Nabiswa & Godfrey, 1994). Isenalumhe & Oviawe, (1988) observed a tendency to prescribe expensive drugs in place of cheaper alternatives in Nigeria. Suprapen (a trade name for a combination of amoxycillin and penicillin V), costing three times more, was prescribed in place of penicillin V in 77.5% of 196 cases. Sixty-eight percent of the prescriptions were judged to have been used unnecessarily. Furthermore, 7 out of 11 physicians who knew that the brand name antibiotic was more expensive than the cheaper available alternative prescribed it all the same. In prescriptions for 299 children with frequent stools, it was common for each one to contain an intestinal antibiotic, a systemic antibiotic, intestinal sedatives, an antipyretic, an antimalarial, some rehydration therapy, and some other drugs, all at the same time. 20 Literature review/2 The practice of inappropriate prescription of drugs is not limited to developing countries. In a District General Hospital in London in 1978, the choice of antibiotics in 19% of 309 courses prescribed therapeutically was considered to be inappropriate. Of 55 antibiotic courses for conventional surgical prophylaxis, only 7% fulfilled all four criteria for appropriate use; 22% fulfilled none (Moss et al., 1981). Similarly, in a Bristol hospital, 28% of 287 prescriptions for antibiotics in 1979 and 35% of 150 prescriptions in 1980 were judged to be unnecessary (Swindell et al., 1983). In a third example, tetracyclines were inappropriately used in children under 8 years of age in the USA. Of 58,639 such children involved in a Medicaid program during a 2-year period, 4,026 (7%) received 7,046 tetracycline prescriptions (Ray et al., 1977). 2.4 Factors influencing drug prescribing Hemminki (1988) distinguishes between factors that affect prescribing at the macro­ level (conditioning factors) and factors that affect individual clinicians. The main conditioning factors are the traditions and education of the population, which may mould both the expectations of the patients and the views of the physicians; medical teaching and professional thinking which determine therapeutic skills; the level and distribution of wealth as well as the political system of the country which affect the importation, regulation and availability of both professional care and drug supply; and the power of the pharmaceuticals industry. The major factors which influence individual prescribers at the micro-level are the demands and expectations of pressure groups and society; the influence of the pharmaceutical industry and research results; and the control measures and regulations imposed by the health authorities. The interrelationships between these factors are illustrated in Fig. 2. Various factors come together to influence the prescribing style of a prescriber making it difficult to assess the role of individual factors. 21 Fig. 2 Factors influencing prescribing behaviour Con tro l and regulations from health authorities, health Insurance systems, medicines committees_______ T Consultants and colleagues Dem ands and expectations of pressure groups and society at large Advertising (Information) O ther health professionals Undergraduate Medical and and postgraduate scientific education journals Physicians's personal characteristics Physician's w orking conditions and therapeutic resources Prescription ^ > Fgedb<,ek Patient's Patient's personal demands and characteristics expectations Traditicm and role of physician's profession Feedback O utcom e of therapy Source: Haaijer-Ruskamp & Hemminki (1993) 21a Literature review/2 The factors that influence prescribing behaviour can also be categorised into four major groups - clinical (prescriber-related), pharmaceutical, laboratory and sociocultural factors. The prescriber-related factors include the prescribing authority given to physicians, which being the hallmark of the profession, sometimes causes them to resist changing their prescribing behaviour even in the face of compelling evidence; a resistance which is fuelled by the fact that patients may be cured from immune response even when drugs are used in a sub-optimal manner (Williams, 1986); the influence from medical colleagues (Linn & Davis, 1972); the training received by the prescriber which may be deficient in practical therapeutics (Hemminki, 1975; Vries de et al., 1995), basic microbiology (Kunin et al., 1987) or the use of therapeutic guidelines, antibiotic guidelines and essential drug lists (Williams, 1986); the availability of objective therapeutic information available to the prescriber (Rainhom & Sangare, 1992); and the heavy workload which means prescribers have only a few minutes to make clinical decisions (Kunin et al., 1987). Mnyika and Killewo (1991) found that recently qualified clinicians, better trained personnel like medical assistants (compared with rural medical aides) and clinicians with access to EDP treatment schedules prescribed more rationally. Health workers who had attended a refresher course were nine times more likely to make an accurate diagnosis and prescribe correctly than those who had not attended. The major pharmaceutical factors which influence prescribing behaviour are the promotional campaigns of the pharmaceutical industry who exert pressure directly through their travelling drug representatives, provision of journals or other printed material, drug samples, drug discounts, organisation or sponsorship of scientific meetings and parties and the largesse in the dispensing of free gifts in the form of pens, diaries, calendars, notepads, clocks etc. (Haaijer-Ruskamp & Hemminki, 1993); the policies and legislation influencing the import, regulation, efficacy, quality, prescription and sale of drugs; the existence of a national drug programme; the availability of drugs 22 Literature review/2 drugs in health facility dispensaries or local chemists and the existence of fake and expired drugs. Prescribers are also influenced by the availability, cost and quality of services of a laboratory, radiography or other diagnostic facilities. Drug companies may supply free antibiotic discs to laboratories for sensitivity testing and thereby promote their own products. In the absence of diagnostic facilities, clinicians have to resort to empirical ("best guess") treatment which may not be appropriate or prescribe multiple drugs to cover all possible ailments (Isenalumhe & Oviawe, 1988). The influence of sociocultural factors on prescribing behaviour is widely recognised. The prescriber may be influenced by both his own cultural preferences and experiences with certain drugs and the prevailing local beliefs about medications. These cultural beliefs include the notion that there is a pill for every ill, the perception that antibiotics are "wonder drugs" which are capable of healing a wide variety of illnesses, the equation of drug effectiveness with the rapidity of response; the perception that expensive drugs are more effective than cheaper alternatives and the superiority of injections over pills or syrup (Kunin et al., 1987). The following excerpts from focus group discussions in the Ashanti-Akim district of Ghana illustrate these beliefs (Waddington & Enyimayew, 1989): "The doctor gave me 6 small pills. I told the doctor that the medicine was too small." (farmer) "One day, a certain man came and asked me where the nurse who gave injections was. I asked him if it was an injection he wanted He answered 'yes' and said, 'you don't give us injections here at all. We don't like taking drugs. What we like are injections.'" (nurse) In Tanzania, it is not uncommon for health workers to be asked by patients to prescribe an injection even if it is not indicated. In fact, some patients walk into clinics 23 Literature review/2 with their own supply of injectables (usually chloroquine or procaine penicillin forte) and with syringe and needles and ask for these to be prescribed (Massele et al., 1993; Mnyika & Killewo, 1991). 2.5 Strategies to improve prescribing behaviour: The Essential Drugs Programme The current strategies employed by countries to promote the rational use of drugs include the formulation of policies and legislation covering all aspects of drug development, manufacture, import, procurement, registration, licensing, prescription, sale, distribution and use of drugs; improvement of educational programmes and research. A number of central bodies are usually formed to act as the regulatory authority for pharmaceuticals and to be responsible for executing the national drug policy. Many countries are increasingly embracing WHO's essential drugs concept and are establishing their own Essential Drug Programmes. WHO (1988c) estimates that up to half of the world's population have little or no regular access to most needed medicines. It therefore behoves developing countries particularly, to use their limited financial resources optimally to ensure available drugs are restricted to those of proven therapeutic effectiveness, acceptable safety and ability to satisfy the health needs of the population. Hence, the selected drugs, called "essential" drugs are those priority drugs that are able to satisfy the health care needs of the majority of the population and should therefore be available at all times in adequate amounts and in appropriate dosage forms (WHO, 1977). In 1977, a WHO committee of experts recommended its first model list of some 200 essential drugs and vaccines to serve as a basis for the development of national drug lists (WHO, 1977). Countries were encouraged to establish their own lists of essential 24 Literature review/2 drugs based on the best available scientific information and relevant to local needs and requirements. These lists were to be regularly updated in the light of advances in drug therapy and other clinical experiences. It was emphasised that exclusion from the list did not imply rejection or that no other drugs were useful but simply that in a given situation the selected drugs were the most needed for the health care of the majority of the population. The WHO Expert Committee recommended that the selection of essential drugs should be accompanied by a concomitant effort in education, training and information of health personnel in the proper use of drugs. The WHO Action Programme on Essential Drugs established in 1981 has provided technical and financial support to developing countries seeking to improve the availability and rational use of drugs. There are now over 120 countries who have national lists of essential drugs and another 60 are developing and implementing comprehensive national drug policies (Anonymous, 1995). The WHO model list of essential drugs is now in its eighth revision and includes some 280 drugs (WHO, 1995). The importance of EDPs in improving rational prescribing is seen in the report of a study in Yemen where a combination of training and provision of an essential drug list (EDL) led to a lower proportion of patients receiving injectables or antibiotics, and a lower average number of drugs per prescription in an EDP area compared with a control area which lacked an EDP (Hogerzeil et al., 1989). The Essential Drugs List for Zimbabwe (EDLIZ), which is one of the most well- established in Africa was first published in 1985 after a revision of an earlier list of 1981. Although Zimbabwe had about 2000 registered drugs, EDLIZ contained only 375 of them. A Zimbabwe Essential Drugs Action Programme (ZEDAP), set up in 1987 with the joint collaboration of the government, WHO and the Danish International Development Agency (DANIDA) drew up a holistic programme to 25 Literature review/2 incorporate the list in a national policy which addressed crucial issues relating to all the stages of the drug supply chain. EDLIZ is now in its third edition and includes 592 drugs; 307 of these drugs are used in specialist facilities and are therefore not normally stocked at the government medical stores (MOH & CW, Zimbabwe, 1994). Other successes of ZEDAP have been the accurate quantification of drug needs, improved availability of essential drugs, the monitoring of the programme indicators using a modified list of WHO indicators and the development and extensive use of training materials with which 160 workshops trained over 6000 health workers. ZEDAP hopes to achieve its target of 85% essential drug availability having already achieved a 73% availability in 1993 and to have efficient government medical stores and a sustainable drug supply management system by 1996 (Chidarikire, 1995; WHO, 1988c). 2.5.1 Essential Drugs Programme Activities in Ghana Ghana has no officially formulated National Drugs Policy although most of the elements inherent in such policy are already present in some form. Some of these elements in existence are the registration system entrusted to a Food and Drugs Board, an Essential Drugs philosophy, a system for regulating the pharmaceutical profession by the Pharmacy Council, and a Cash and Carry cost recovery scheme designed to improve drug financing. The Essential Drugs Programme of Ghana has not yet been established although here too, elements of the programme already exist. For example, a committee responsible for preparing and revising the EDL has been in place for a number of years. Various consultants from the WHO and Netherlands in collaboration with the National Drugs Committee and the Policy, Planning, Monitoring and Evaluation Division of MOH have been involved in drawing up Ghana's EDP and the draft document has reached an 26 Literature review/2 advanced stage. Donors including the Overseas Development Agency and DANLDA have expressed interest in supporting the programme (D Asiama, personal communication). With regard to the history of the EDL, a national formulary was first issued in 1968; later in 1973, an approved list of drugs was prepared. A list of non-essential drugs on the Ghana market was also produced at the time. However, the first Provisional Essential Drugs List and National Formulary (EDL/NF) based on the WHO model was formally launched in 1988. A National Drugs Committee, composed of 11 members was set up by MOH and charged with the regular revision of the EDL/NF. The adoption of the EDL was part of a national policy to ensure that all Ghanaians have access to effective, safe and affordable drugs of good quality, in both the public and private sectors and that these are rationally used. It was expected that the initiative would reduce or even eliminate the problem of chronic shortage of drugs within the public sector. The problem had been one of ineffective drug management from MOH headquarters downwards and irrational prescribing. Having defined the most essential drugs needed, efforts were made to ensure their regular supply. A physical inventory of all pharmacies and dispensaries in the public sector was undertaken by the Faculty of Pharmacy of the University of Science and Technology, Kumasi with World Bank support. Following the inventory, a task force was set up by MOH to estimate annually the drug requirements of the public health system, using morbidity data and allowing for contingencies. Although it was intended that the EDL/NF be revised every two years, it was not until 1993 that the next edition was released (MOH, 1993). There have been a number of problems associated with the revised manual. Financial problems have delayed the printing of sufficient copies of the manual. In fact, it is only with a recent US AID-support that copies of the manual are now being printed. 27 Literature review/2 In 1994, the National Drug Committee, with assistance from some consultants from the Netherlands was charged to study the pharmaceutical sector and to draw up a plan of action to improve its functions. They reported that despite its substantial impact on procurement in the public sector, the revised EDL/NF was little known in the country and did not appear to have had much impact on prescribing (National Drugs Committee, 1994). This was also the case with the 1988 manual where a partial assessment by the RPM study indicated its availability in only 45% of public health facilities (Rankin et al., 1993). The revised manual has not been actively marketed; neither has there been a national programme to promote rational drug prescribing (or to enforce compliance with the EDL) as recommended by the WHO Expert Committee on the Selection of Essential Drugs (WHO, 1977). Interestingly, the EDL has no legal backing. The manual provides guidelines in the management of a number of clinical conditions and a few selected infectious diseases. Commonly encountered diseases such as enteric fever, septicaemia, pelvic inflammatory diseases and cellulitis are not included. Reserve antimicrobial drugs for common infections are not mentioned. Whereas the manual recommends cotrimoxazole or amoxycillin for lower urinary tract infections (UTI) and procaine penicillin or amoxycillin for gonorrhoea, recent studies indicate that over 75% of common urinary isolates and over 90% of Neisseria gonorrhoeae are resistant to these drugs. In contrast, the same studies report a 90-98% sensitivity to such reserve antimicrobials as norfloxacin, cefuroxime and ceftriaxone (Adu-Sarkodie, 1993/94; Newman, 1990). It has been suggested that the latter agents be included in the EDL/NF and efforts made to guard their indiscriminate use to prevent the early development of resistance (Adu-Sarkodie, 1995). 28 Literature review/2 In comparison, the treatment protocol for gonorrhoea in the EDLIZ provides for first, second and third line drugs which include such reserve antimicrobials as kanamycin, norfloxacin, spectinomycin and ciprofloxacin (MOH & CW Zimbabwe, 1994). The WHO model list also provides for the inclusion of reserve antimicrobials in view of the increasing prevalence of resistant bacteria (WHO, 1992a). Fortunately, in preparation for the next revision of the EDL/NF, a circular inviting comments and suggestions on the list has been issued to health providers throughout the country (National Drugs Committee, 1995). It is hoped that these recommendations will be examined in the preparation of the next manual. The national strategies to promote rational use of drugs within the framework of an EDP have only just been published (MOH, 1995). The strategies are listed as follows: • standard treatment protocols will be developed and both the public and private sector prescribers will be trained in their proper use • regulations will be made to limit the range of drugs to be prescribed according to category of worker and level of operation • measures will be taken to encourage the private pharmaceutical sector and the medical profession to utilise the EDL through wide distribution of the document and a combination of education campaigns and enforcement as required. It would appear the MOH has recognised some of the weaknesses of the EDL and has identified appropriate strategies to address them. However, the details of the plan for strategy implementation were not provided e.g. how enforcement of prescribing regulations or compliance with the EDL will be achieved. It is commendable that apart from the national EDL/NF, some hospitals have produced their own manuals to guide medical students, medical assistants, nursing staff and doctors in the management of common medical problems. The Komfo 29 Literature review/2 Anokye Teaching Hospital has two such manuals for use in the Paediatric and Medical wards (Fleming, 1991; Mackie & Asafo-Agyei, 1990). 2.6 Methodological review There are two major types of research employed in investigating prescribing patterns: qualitative research (sometimes also called anthropological) and quantitative or biomedical research. In the qualitative approach, the individual perspective is emphasised and there is therefore an extensive interaction with the study population e.g. physicians or pharmacists in health facilities. This approach is especially suitable for studying the meaning of drug use, such as the psychosocial and cultural basis for prescribing behaviour. It could be used to compare prescribing practices in different institutions for example, in terms of staff attitudes, patient demands and the quality of interaction between prescribers and patients. The most frequently used methods of data collection are participant observation and in-depth, open-ended interviews. These techniques are often time-consuming and not appropriate for evaluating the extent of irrational prescribing in different situations, for which the quantitative approach is needed. In the quantitative approach, the investigator is able to measure various prescribing indicators which can be subjected to statistical analysis. These indicators may be analysed for different types of prescribers or facilities and compared with results of other studies. The measurement of the indicators can also provide the basis for evaluating any interventions to improve prescribing. The quantitative approach has the disadvantage of being oversimplified; it ignores the sociocultural aspects of drug prescribing. Clearly then, the quantitative methodology and the qualitative one 30 Literature review/2 compliment each other. The most widely used data collection method in the quantitative approach is the structured interview. Studies on prescribing patterns may be based on either prospective (data from current patients as they present for consultation) or retrospective data (i.e. past medical records). The former method has the advantage of presenting complete data but suffers from the bias of a change in prescribing practices when prescribers are aware that their behaviour is being observed. Other disadvantages pertain to seasonal variations in morbidity patterns, peculiarities in staffing and inconsistencies in drugs supply, all of which influence prescribing patterns. Collection of retrospective data eliminates most of these biases but has the disadvantage of presenting incomplete data where records are not available or well-kept. One of the most popular methods of investigating prescribing patterns has been described by EMRUD and WHO. A basic core of prescribing indicators (e.g. average percentage of drugs prescribed at each visit and percentage of patients receiving an injection) which can be studied in a standard way in different areas at different times is presented. Complementary indicators to some specific indicators are also presented. Various studies have generally not found any significant differences in prescribing indicators (with the exception of the percentage of patients receiving injection) obtained from retrospective and prospective data (Massele et al., 1993; WHO, 1993). The WHO drug use indicators have been used to identify priority areas for action. In Zimbabwe where 94% of drugs are prescribed by their generic names, this feature is not a priority for its EDP whereas in Ecuador, where only 37% generic prescribing, there is a clear need to allocate resources to this area (WHO, 1993; Hogerzeil et al., 1993). The indicators have also been used to quantify the impact of the EDP on rational prescribing in Yemen (Hogerzeil et al., 1989). A number of the WHO drug use indicators are investigated in the present study. 31 Literature review/2 The WHO/INRUD methodology was used in the RPM assessment of drug utilisation in Ghana (Rankin et a l, 1993). Twenty health facilities comprising were selected by multi-stage sampling (see section 1.2). Although, the study was originally intended to cover all 10 regions in the country, only 5 were covered owing to managerial problems that prevented the training of all the regional Cash and Carry Coordinators who were to be involved in the study. Within a selected region, a district was randomly selected. Then 2 health centres were selected randomly from all health centres located within the selected district. This selection procedure was followed for logisitical reasons as the assessment team also assessed the stock management, procurement and pricing practices in the Cash and Carry system at the various levels. However, this procedure carried with it, a potential selection bias in that the prescribing practices in health centres within the one district may have differed systematically (with regard to such influences as availability of drugs, recent training of prescribers in treatment guidelines etc) from those of the other health centres in the region. A random selection of the 2 health centres from a list of all the health centres within each selected region would have eliminated the bias and made a generalisation of the findings more justifiable. 2.7 Conclusion The irrational prescribing of drugs is a problem in many developing countries. The situation is compounded by the fact that most of these countries lack effective policies relating to all aspects of drug supply. Many countries have adopted an Essential Drugs Programme in an effort to improve both the availability and the rational prescribing of drugs. Such programmes have to be accompanied by educational and training programmes if they are to achieve their objectives. 32 CHAPTER THREE 3.0 METHODOLOGY 3.1 Study type The study was cross-sectional comparative in type. Data on the prescribing patterns in different health facilities in the Wassa West district were collected over a short period of time. The prescribing patterns in the different health facilities were also compared with one another. 3.2 Study period Data on prescribing indicators were collected over a four week period from October to November, 1995. 3.3 Operational definition of terms Prescribing refers to a formal practice in which authorised clinicians write a list of selected drugs to be used for the cure, alleviation or prevention of a clinical condition. The Medical and Dental Decree, 1972 (NRCD 91) limits prescribing of drugs to doctors, medical assistants and midwives. Antibiotics refer to antibacterial drugs, irrespective of whether they are naturally occurring or synthetically produced that are used for the treatment or prevention of a bacterial infection. In conformity with the classification in the latest WHO Model List (WHO, 1995), topical antibiotics and metronidazole (which is also antiprotozoal) were counted as antibiotics. Antituberculous and antileprosy drugs were not included in the 33 Methodology/3 definition except when they were used singly for non-tuberculous or non-leprosy conditions e.g. rifampicin for meningococcal meningitis. Generic names refer to the non-proprietary names rather than the brand names by which drugs are prescribed. Names of drugs were considered generic if they corresponded to those in the EDL/NF (MOH, 1993). Injectable drugs refer to those drugs that are administered by the intramuscular, intravenous, intradermal, subcutaneous, subconjunctival or intrathecal routes. Essential drugs list is a list of drugs that are most needed to satisfy the health needs of the majority of the population and should therefore be available at all times in adequate amounts and in appropriate dosage forms (WHO, 1977). i Retrospective data describe prescribing patterns during patient encounters that took place in the past. Prospective data describe prescribing patterns during patient encounters that take place on the day of the survey. 3.4 Definition and measurement of variables studied The definition and indicators of the variables studied are shown in table 3.1. i 3.5 Study population The study population comprised: • all government health institutions in the Wassa West district • all outpatient clinical record cards between June 1994 and June 1995 34 Methodology/3 • all outpatient prescription forms • all inpatient treatment records on a specific day of the week • all Heads of government health institutions in the district Tab le 3.1 D e fin it io n and m ea su re m e n t o f se lec ted va ria b le s Variable Definition Indicator Average number of drugs The number of drugs prescribed for each patient. Sequential regimen e.g. Inj. Chloroquine followed by oral Chloroquine was counted as one drug The total number of drug items prescribed divided by the total number of patients for whom they were prescribed Percentage of drugs prescribed from the essential drugs list The proportion of drugs prescribed which are listed on the 1993 Essential Drugs List of Ghana The total number of drugs prescribed which are on the EDL divided by the total number of drugs prescribed, multiplied by 100 Percentage of drugs prescribed by generic name The proportion of drugs prescribed by generic names as contained in the EDL of Ghana The number of drugs prescribed by generic name divided by the total number of drugs prescribed, multiplied by 100 Percentage of patients receiving an injectable drug The proportion of patients for whom injectable drugs are prescribed The total number of patients for whom at least one injectable drug is prescribed divided by the total number of patients surveyed, multiplied by 100 Average number of injectable drugs The number of different injectable drugs prescribed per patient The total number of injectable drugs prescribed divided by number of patients receiving at least one injectable drug Average number of injection episodes The number of times each patient is required to receive an administered injectable drug The total number of times an injectable drug is prescribed to be administered divided by the total number of patients receiving at least one injectable drug Percentage of patients receiving an antibiotic The proportion of patients for whom antibiotics are prescribed The total number of patients receiving at least one antibiotic divided by total number of patients surveyed, multiplied by 100 Average number of antibiotics The number of different antibiotics prescribed per patient The total number of antibiotics prescribed divided by the total number of patients receiving at least one antibiotic Percentage of total drug cost due to antibiotics Average cost of drugs as obtained from the district hospital or two private chemical shops in Tarkwa The total cost of antibiotics divided by total cost of all drugs multiplied by 100 Appropriateness of prescriptions For a given clinical condition, the choice of prescribed drug, dose, regimen, route of administration, duration of treatment, to the extent possible are considered inappropriate The observer certifies that all criteria have been met after analysis of prescriptions Type of prescriber The rank of prescribers at the health facility on the day of visit The Head of the facility mentions the rank of prescribers in her faciity Availability of EDL A copy of the1993 edition of the manual "Essential Drugs List and National Formulary of Ghana with Therapeutic Guidelines" Is available at the health institution The manual is made available for inspection Availability of drugs Drugs used for the treatment of common conditions are available at the health facility visited The drugs for treatment of conditions are seen on inspection on the day of visit to health facility 35 Methodology/3 3.6* Sampling 3.6.1 Sampling method Government health facilities A census o f all governm ent health facilities in the W assa W est district w as taken. All the facilities w ere included in the study. Outpatient clinical records (retrospective data) Th6 O utpatients' D epartm ent (OPD ) Register contains a chronological listing o f all patients attending the health facility for daily consultations with their unique identity numbers. Separate registers are kept for new patients and reattendants. Using the OPD register, a sampling frame o f new outpatients attending the health facility between June 1, 1994 and June 30, 1995 was obtained. The patient encounters to be studied w ere selected by system atic sampling. The first patient encounter was selected randomly using a random digital table from the list o f all patients betw een the first patient recorded in the register and the patient w ith a serial number corresponding to the sampling interval. The identity num bers o f selected patients w ere recorded from the register and their O PD clinical record cards retrieved. Prescription forms (prospective data) Prospective data from prescription forms were collected by systematic sampling o f all prescriptions that w ere presented at the dispensary o f the District Hospital. The 36 Methodology/3 collection of prospective data at this facility was necessitated by a relatively large proportion of incomplete data obtained from the retrospective records. Inpatient treatment records The treatment records of all patients on admission at the District Hospital on a randomly selected day of the week were studied. Heads o f health facilities All the Heads of the government health facilities were identified and interviewed. 3.6.2 Sample size Government health facilities The study of prescribing patterns covered all 7 government health facilities in the Wassa West district namely: 1. Tarkwa Government Hospital (TGH) 2. Nsuaem Health Centre (NHC) 3. Bogoso Health Centre (BHC) 4. Himan Health Centre (HHC) 5. Huni Valley Health Centre (HVHC) 6. Dompim Health Centre (DHC) 7. Simpa Health Centre (SHC) 37 Methodology/3 Retrospective patient encounters The determination of sample size was in accordance with the recommendations of the WHO/INRUD methodology which has been well tested in over 14 countries (Hogerzeil et al., 1993; WHO, 1993). The guiding principles for the sample size determination were as follows: • individual prescribers tend to exhibit consistent practices over time, so that a sample drawn at one point in time will provide basically the same results as a sample that covers a longer period; • above a certain number of encounters, adding additional encounters to a sample within a facility adds very little new information. For comparing prescribing practices between individual health facilities, the minimum number per facility is 100. For a study in which prescribing encounters from several (commonly 20) health facilities are collected and studied as a whole, the minimum number is 30 per facility (See Annex 1); • the above sample size of 100 measures facility-specific percentage indicators with a 95% confidence interval of plus or minus 10%. On the basis of these guidelines, one hundred retrospective encounters at each of the seven health facilities in the district were selected for study. A total of 700 outpatient client cards were thus obtained for study. The sample size agrees with that obtained by calculation on the STATCALC Epi Info computer software (Table 3.2). T a b le 3 .2 S a m p le s iz e c a lc u la tio n fo r a s in g le fa c il ity Largest study population in a single facility 1 7 ,500 E xpected p revalen ce o f antibiotic prescribing 5 0 .0 % W o rst acceptab le frequency 6 0 .0 % Confidence Level 80% 41 90% 6 7 95% 96 99% 164 38 Methodology/3 The sample size for a single facility at the 95% confidence interval of 10% from the above calculation is 96. If prescribing practices of a prescriber in a single facility were independent of each other, multiplying 96 by 7 will give a total of 672 encounters to be selected from all the health facilities covered in the study. A smaller sample size than that suggested by calculation is actually needed as prescribing practices tend to be consistent within a facility. Even so, a sample size of 700 was selected after rounding off and also to accommodate any encounters that may need to be rejected. Prescription forms Prospective data from 100 prescription forms were obtained by systematic sampling from a daily average of 60 prescriptions presented at the dispensary of the Tarkwa Government Hospital over a 5-day period in November, 1995. Inpatient treatment records The treatment records of 45 inpatients (representing about 65,2% bed occupancy) at the Tarkwa Government Hospital on a randomly selected day of the week were examined. Heads o f government health facilities All 7 Heads of the government health facilities covered in the study were interviewed about selected health facility indicators e.g. the availability of the national EDL manual and the availability of drugs for treating common diseases. 39 Methodology/3 3.7 Data collection 3.7.1 Permission to proceed Permission to undertake the study was obtained from the following: School of Public Health (Director and Academic Supervisors) The Regional Director of Health Services, Western Region The District Chief Executive, Wassa West District The Director of Health Services, Wassa West District (also the Field Supervisor for the study) The Heads of health facilities covered in the study The School of Public Health had previously obtained the approval of the relevant authorities for the district to be used as a Field Site for a 17-week field residency programme for residents of the school. Part of the field programme was used for data collection for resident's dissertation. 3.7.2 Data collection tools and techniques The tools and techniques employed for the study are shown in table 3.3: Tab le 3.3 Data c o lle c tio n to o ls and te c h n iq u e s V a ria b le /In d ic a to r Data c o lle c tio n to o l Data c o lle c tio n tec h n iq u e S o u rc e o f d a ta Prescribing indicators e.g . % g eneric drugs, % patients receiving antib iotics etc. Q uestionnaire Using ava ilab le inform ation Q P D clinical records; Prescription form s; Inpatient trea tm e n t records A vailab ility o f EDL; drugs fo r trea tm e n t o f com m on d iseases Q uestionnaire; Checklist Interview ; Inspection H ealth facility H eads; Consulting room; d ispensary T yp e o f prescribers Q uestionnaire In terview H ealth facility H ead s 40 Methodology/3 Consultation data at the health facilities are recorded on the OPD clinical record card. The card provides data on the patient's name, age, sex, address and identity number; name of facility; date of visit; clinical history and findings; diagnosis, treatment and prescription. With the exception of the history and examination findings, all the other data were collected and entered directly on a questionnaire form (Annex 2) designed on the Epi Info 6.03 computer programme (Center for Disease Control and Prevention, Atlanta; World Health Organisation, Geneva). Some of the key data with multiple responses e.g. names of drugs, names of antibiotics etc. were entered again in related files, thereby allowing consistency checks to be made to improve the quality of data. Individual cards were coded both manually and electronically for their serial numbers and name of facility. Data from prescription forms and inpatient records were similarly collected. Data on non-drag items in the treatment plan such as blood transfusion or suturing were excluded. Information on the availability of the EDL manual and drags for the treatment of common diseases was supplied by the Heads of the health facilities. Their responses were ascertained by inspection of the available manual and drags. The results of the study were presented to prescribers and district health authorities at a meeting at the end of the study. Some reasons for the observed prescribing patterns were obtained from the prescribers at this meeting. 3.7.3 Pre-test Pre-testing of the data collection questionnaire forms was undertaken at the Cape Coast Central Hospital. Data relating to the prescribing indicators were analysed and summarised in order to assess the suitability of the questionnaire design. The pre-test enabled a few changes to be made; for example the list of possible diagnoses was expanded and the variable on the number of injection episodes was added to the 41 Methodology/3 original form. Pre-testing also served to improve the data entry skills of the investigator. At the Tarkwa District Hospital, the availability of the study population was determined. While it was originally intended to collect retrospective data from prescription forms, it became clear that these were not available as patients demanded them to facilitate their drug expenditure refund claims. 3.8 Data processing and analysis The data were analysed separately for individual health facilities and then collectively for all of them using the Epi Info version 6.03 computer software. Data were summarised by tables and graphs; relevant statistical tests were performed where applicable. Data on indications of antibiotic use were analysed manually as this could not be done with the computer software used in the study. 3.9 Advantages and limitations of methodology 1. The selection of health facilities was limited to the government facilities for logistical reasons and also for the more direct control over the activities of these facilities by the DHMT. Recommendations could be more easily implemented and monitored. However, because prescribing patterns and influences at the selected facilities may differ from those of the other health facilities e.g. Mines Hospital or private clinics, findings from the study may not generalised to the latter. 2. The time frame for the retrospective study was limited to one year in order to accommodate any seasonal variations in prescribing patterns or other influences such as inconsistencies in drug supply. In order to overcome the 42 Methodology/3 problem of incomplete data encountered from retrospective records at the District Hospital, prescribing indicators were also determined from prospective data. 3. The collection of prospective data from prescription forms is subject to weaknesses such as a change in prescribing practices arising from prescribers' awareness of the study, poor random selection because of the lack of a sampling frame and a selection bias whereby the data collector favours the selection of certain prescriptions. In order to overcome some of these weaknesses, prescribers were not made aware of the study. Also, the dispensing officer who collected the data was given clear instructions on how to systematically collect them; he was also not informed of the parameters under investigation and was supervised intermittently. 4. The prescribing patterns were not analysed in relation to the specific ranks of prescribers. Being a preliminary study, it was considered more appropriate to first measure prescribing patterns at the macro-level (i.e. health facility). Follow-up investigations could then be dictated by observed patterns. 5. The appropriateness of treatment was assessed with respect to the prescribing indicators and to a lesser extent, the actual drugs selected. However, given the common practice of poor documentation, it is possible that treatment may have been given for a genuinely diagnosed medical condition that had not been recorded. Naturally, assessment of appropriateness could only be based on what had been recorded on the outpatient card. 6. The study recognised that merely issuing a prescription was no guarantee that the patient will obtain all the prescribed drugs or that even when he obtained them, he will comply with the treatment. 43 Methodology/3 3.10 Ethical considerations In spite of a probable notion that an "external auditor" had come into their district to as it were point out faults in their prescribing practices, the prescribers in the Wassa West district were most cooperative. This was evident in the assistance they provided with the data collection. Their enthusiasm was again in evidence when they attended and actively participated in a meeting at which findings of the study were presented. The explanation that the study did not intend to pass a professional or moral judgement on individual prescribers helped to secure their cooperation. 44 CHAPTER FOUR 4.0 RESULTS 4.1 Demographic data The age and sex distribution of 700 OPD patients from the retrospective encounters is presented in table 4.1. T ab le 4.1 Age and sex d is tr ib u t io n o f p a tie n ts a tte n d in g g o ve rn m e n t hea lth fa c ilit ie s Age (Years) Female Male Total Percent <1 40 41 81 11.6% 1 - 4 84 77 161 23.0% 5 -14 46 54 100 14.3% 1 5 -4 4 170 126 296 42.3% 4 5 -5 9 25 21 46 6,6% 60+ 11 5 16 2.3% T o ta l 376 324 700 100,0% The ages of the patients ranged from 1 week to 80 years with a mean of 17.9 years. About 51.2% of patients were aged 15 years or older and 53.7% were female. For the one day prescribing practices survey at the Tarkwa Government Hospital, the ages of the in-patients ranged from 1 month to 68 years with a mean of 25.1 years. T\Menty-six (57.8%) of the 45 patients on admission were females. 4.2 Diseases encountered Table 4.2 lists the frequency of diseases encountered at the OPD of the government health facilities. The total number of diseases in excess of the number of patients reflects patients with multiple diagnosis. 45 Results/4 Tab le 4.2 D iseases encoun te red in 700 O PD c lin ic a l reco rd s D isease D is tr ic t H osp ita l 6 H ealth C entres T o ta l P e rc e n t Accidents 3 7 10 1,2 Acute eye Infection 3 4 7 0.8 Anaemia 1 6 7 0.8 Bites and minor trauma 6 35 41 5.0 Boils and wound infections 1 23 24 2.9 Chickenpox - 5 5 0.6 Diarrhoeal diseases 6 56 62 7.5 Dysentery 2 6 8 1.0 Ear infection 1 5 6 0.7 Gastrointestinal disorders 4 8 12 1.5 Gynaecological disorders 1 4 5 0.6 Hypertension - 6 6 0.7 Intestinal worms 2 25 27 3.3 Jaundice and viral hepatitis 1 4 5 0.6 Liver abscess 1 - 1 0.1 Malaria 66 341 407 49.3 Malnutrition - 4 4 0.5 Measles - 9 9 1.1 Medically fit 2 - 2 0.2 Mental disorder 1 1 0.1 Pelvic inflammatory disease 1 3 4 0.5 Pneumonia - 2 2 0.2 Pregnancy complications 2 6 8 1.0 Pyr