Received: 7 August 2018  |  Revised: 28 August 2018  |  Accepted: 16 November 2018  |  First published online: 3 December 2018 DOI: 10.1002/ijgo.12717 C L I N I C A L A R T I C L E G y n e c o l o g y Availability and prescription of misoprostol for medical abortion in community pharmacies and associated factors in Accra, Ghana John K. Ganle1,2,* | Nana T. Busia1 | Ernest Maya1 1Department of Population, Family and Reproductive Health, School of Public Abstract Health, University of Ghana, Legon, Accra, Objective: To assess misoprostol availability at community pharmacies and determine Ghana 2 factors affecting misoprostol prescription for medical abortion.Stellenbosch Institute for Advanced Study, Stellenbosch University, Stellenbosch, Methods: A cross- sectional quantitative survey of randomly sampled community South Africa pharmacies and their corresponding pharmacists/pharmacy workers was conducted in *Correspondence the city of Accra, Ghana. Structured questionnaires were administered to collect data John K. Ganle, Department of Population, between May 1 and July 28, 2016. Descriptive statistics (frequencies and proportions) Family and Reproductive Health, School of Public Health, University of Ghana, Legon, and bivariate and logistic regression analysis were used to analyze the data. Accra, Ghana. Results: Of the 165 community pharmacies surveyed, approximately half (83 [50.3%]) Email: jganle@ug.edu.gh stocked misoprostol. Availability of misoprostol however decreased when moving from first class to third class residential areas. 44 (26.7%) of the respondents had pre- scribed the drug at some time for medical abortion, but 140 (84.6%) indicated they would not prescribe the drug for medical abortion in future. Factors that significantly predicted misoprostol prescription for medical abortion included sex of the pharma- cist/pharmacy worker, demand, and availability of misoprostol. Conclusions: Demand for misoprostol for medical abortion was found to be high but only half of community pharmacies stocked it, and most pharmacy workers did not wish to prescribe the drug. If community pharmacies are to effectively contribute to expanding access to safe abortion services in Ghana, the disparity between misopros- tol provision and demand needs to be addressed. K E Y W O R D S Availability; Community pharmacies; Ghana; Medical abortion; Misoprostol; Prescription 1  | INTRODUCTION 5.3 million women also suffer temporary or permanent disability as a result of unsafe abortion.2 Unsafe abortion is a significant public health problem, accounting for Medical abortion offers a vital alternative to surgical abortion for 13% of maternal mortality in low- resource countries.1 WHO defines women with early pregnancies who wish to avoid a surgical proce- unsafe abortion as a procedure for terminating an unplanned preg- dure.3 Presently, three regimens exist for medical abortion in Ghana nancy carried out either by persons lacking the required skills or in an generally, and in the area where the present study was conducted environment that does not conform to minimum medical standards, or specifically: misoprostol alone, methotrexate followed by misopros- both.2 It is estimated that unsafe abortion accounts for 70 000 deaths tol, and, by far the most commonly used method, mifepristone fol- yearly, over 99% of which are in low- income countries.1 An additional lowed by misoprostol.4 Misoprostol, the subject of interest in this Int J Gynecol Obstet 2019; 144: 167–173 wileyonlinelibrary.com/journal/ijgo © 2018 International Federation of  |  167 Gynecology and Obstetrics 168  |     Ganle eT al. paper, is a synthetic analogue of the prostaglandin E1 that entered to conduct business) in the Greater Accra Region at the time of this the global market in the late 1980s.5 Marketed as Cytotec (Pfizer, study.20 Out of this number, 281 (36.4%) were located in the Accra New York, NY, USA), misoprostol is an effective and progressively Metropolis. These were distributed across three main residential areas popular option for self-a dministration, especially in low- resource in the city, which differ in socioeconomic characteristics.21 First class countries.6 Mifepristone, followed by a prostaglandin (such as miso- residential areas are generally well planned, have well developed prostol), is considered the gold standard for medical abortion.7 A infrastructure, and expansive, landscaped properties.21 The richest regimen of 200 mg mifepristone and 800 μg buccal or vaginal miso- segment of society usually occupy these areas. Second class areas are prostol is 95%–98% effective.8 In several settings where mifepris- middle- income areas principally occupied by working class people.21 tone is inaccessible, however, misoprostol is commonly used alone These areas are generally better- planned, albeit in need of infrastruc- for early pregnancy termination.7 More importantly, there is evidence ture services. Third class areas are generally low- income, heavily pop- that women are capable of safely and effectively self-a dministering ulated, and largely unplanned.21 misoprostol either alone or in combination with mifepristone, even in The target population was primarily community pharmacists resource- poor settings.7 located in the three strata of residential areas. However, the regular While it is fairly easy to access safe abortion services in most high- absence of pharmacists in community pharmacies in low- income set- income countries, safe abortion has become a privilege of the rich tings is recognized.13–18 In the absence of the community pharmacist, in settings with highly restrictive induced abortion regulations.9,10 In pharmacy technicians or medicine counter assistants, or other (on the Ghana, maternal mortality is the second highest cause of death among job trained) pharmacy workers were surveyed. women of reproductive age, with approximately 20.8% of maternal Since the number of pharmacies were known, Yamane’s formula deaths resulting from unsafe induced abortions in some major health for calculating sample size from a finite population was used.22 A sim- facilities in the country.11 Notwithstanding the fact that Ghana has a ple random sampling technique was then used to select the requisite relatively liberal abortion law, access to safe, legal induced abortion in sample number from the total of 281 pharmacies. For each of the public health institutions is often limited.11,12 More recently, one pub- pharmacies included, one pharmacist/pharmacy worker was selected lished study12 reported on the availability of medical termination of to take part in the study. In cases where there was only one phar- pregnancy kits in pharmacies in Ghana. The availability of such prod- macist/pharmacy worker, that person was automatically included in ucts in community pharmacies could potentially expand access to safe the study. However, where there were more than one pharmacist/ abortion services in Ghana. pharmacy worker, the most senior (in terms of years of practise) was Several studies have shown that in low- income countries where selected to respond to the questionnaire. legal induced abortion services are not easily accessible, community Structured questionnaires were designed and administered to the pharmacies are frequently the first line healthcare providers when pharmacists/pharmacy workers to collect information on availability of women seek induced abortion services.13–16 In Ghana, community misoprostol in community pharmacies, and factors affecting misopros- pharmacies have been reported to be the first port of call for the tol prescription. The questionnaires were pre- tested in an adjoining majority of Ghanaians seeking healthcare advice or treatment.17,18 municipality not included in the study, and all necessary amendments Despite this recognition, few studies have been carried out to ver- were made to the questionnaires before collection of the study data. ify the availability of misoprostol in community pharmacies and The data were collected by a researcher (NTB) between May 1 and factors affecting its prescription for medical abortion. The present July 28, 2016. The questionnaires were both interviewer- administered study aimed to assess misoprostol availability at community phar- (a researcher [NTB] asked questions and recorded responses) and macies and determine factors affecting misoprostol prescription for self- administered (respondents read the questions, and recorded medical abortion. their responses unsupervised). The researcher personally visited all pharmacies to administer questionnaires or distribute them for self- administration. The researcher also personally revisited all pharmacies 2  | MATERIALS AND METHODS to retrieve self- administered questionnaires. Completed questionnaires were retrieved, checked for complete- The present cross- sectional quantitative survey collected data ness, coded, and the data analyzed using Stata version 13.1 (StataCorp, between May 1 and July 28, 2016, using structured questionnaires. College Station, TX, USA). Descriptive statistics were used to describe The study was conducted among pharmacy workers in the Accra the sociodemographic characteristics of respondents, availability of Metropolis of the Greater Accra Region of Ghana—the most populated misoprostol, and whether pharmacists/pharmacy workers had pre- district in Ghana with an estimated total population of 1 665 086, scribed or would prescribe misoprostol for medical abortion. Bivariate comprising 51.9% females and 48.1% males.19 Ethical clearance was analysis using χ2 test was used to assess the association between obtained from the Ghana Health Service ethical review committee. different characteristics of the respondents and dispensing of miso- Informed written consent was obtained from all respondents. prostol. Logistic regression was fitted to further assess the strength According to the Pharmacy Council of Ghana, there were 771 of association among variables that were significantly associated with registered retail community pharmacies in good standing (i.e. pharma- misoprostol dispensing. Confidence level was held at 95% and P<0.05 cies that have renewed their yearly license and therefore are eligible was considered statistically significant. Ganle eT al.      |  169 3  | RESULTS with pharmacists and “on the job trained” staff (Table 3). However, when asked in a follow- up question whether they would prescribe it to women seeking induced abortion services in the future, the A total of 165 pharmacists/pharmacy workers participated in the pre- majority (140 [84.6%]) said they would not. Pharmacists and “on the sent study, the number arrived at through the sample size calculation, job trained” staff were more likely than other pharmacy workers to all of whom completed a questionnaire. report that they would not prescribe misoprostol for medical abor- A summary of the sociodemographic characteristics of respon- tion. The reasons for the decision to prescribe the drug or not are dents can be found in Table 1. The majority were female (90 [54.6%]), presented in Table S1. The majority of respondents who said they and 118 (71.5%) had tertiary education. With respect to qualifications, would not prescribe it said it was against their religious belief (71/128 69 (41.8%) were pharmacists, while 10 (6.0%) were members of staff [55.5%]); 21/128 (16.4%) said they were not sure of the dose to give “trained on the job”. and 36/128 (28.1%) said they would not prescribe it because it was Table 2 provides information on the availability of, and demand outside their scope of practice. for, misoprostol. Approximately half (83 [50.3%]) of the pharmacies Tables 4 and 5 show the results of bivariate and logistic regression surveyed stocked misoprostol. However, availability of misoprostol analyses investigating the association between a number of indepen- declined from first class (high-i ncome) to third class (low- income) res- dent variables and ever prescribing misoprostol. Using bivariate analysis, idential areas. Weekly demand was highest in third class, followed by only the sex of the pharmacy worker (P=0.005), availability of misopros- second class, residential areas. However, monthly demand was highest tol in the pharmacy (i.e. currently stocks misoprostol) (P<0.001), and in the first class, followed by second class, residential areas. demand for misoprostol (P<0.001) were significantly associated with a When asked whether or not they had ever prescribed misopros- pharmacy worker ever prescribing misoprostol (Table 4). tol for medical abortion, 44 (26.6%) pharmacy workers said they Factors that showed association in the bivariate analysis were fur- had. Pharmacy technicians and medicine counter assistants were ther investigated using logistic regression analysis. This was done by slightly more likely to report ever prescribing misoprostol compared estimating crude and adjusted odds ratios, and the results are shown in Table 5. The odds of a female pharmacy worker ever prescribing TABLE  1 Sociodemographic characteristics of respondents (n=165). misoprostol for medical abortion was 0.48 times lower than a male worker (crude odds ratio [cOR] 0.48, 95% confidence interval [CI] Characteristic No. (%) 0.23–0.97). When other factors (i.e. currently stock misoprostol and Age, y demand for misoprostol) were adjusted for, the odds of a female phar- 20–24 22 (13.3) macy worker having ever dispensed misoprostol compared to males 25–29 57 (34.6) reduced to 0.37 times lower (adjusted odds ratio [aOR] 0.37, 95% CI 30–34 35 (21.2) 0.16–0.85), and the association was no longer statistically significant. 35–39 31 (18.8) Similarly, community pharmacy workers who stocked misoprostol ≥40 20 (12.2) were 4.24 times more likely to indicate they would ever consider dis- Sex pensing misoprostol when compared with those who did not stock Male 75 (45.5) misoprostol (aOR 4.24, 95% CI 1.18–10.0). This association was Female 90 (54.6) statistically significant. Religion Christian 153 (92.7) 4  | DISCUSSION Other 12 (7.3) Marital status The present study was one of only a few to assess misoprostol availabil- Married 71 (43.0) ity in community pharmacies and the factors affecting its prescription Single 88 (53.3) for medical abortion in Ghana. The results showed that misoprostol Other 6 (3.6) was only available in approximately half of community pharmacies in Education Accra, with the drug becoming less available in low- income residen- Tertiary 118 (71.5) tial areas. Weekly demand was highest in third class residential areas, Secondary 47 (28.5) where the drug was less available. However, monthly demand was Qualification highest in first class residential areas where the drug stock was also Pharmacist 69 (41.8) highest. A little over a quarter (44 [26.7%]) of the respondents had ever given the drug to clients seeking induced abortion services, but Pharmacy technician 27 (16.4) the majority (140 [84.6%]) indicated that they would not prescribe the Medicine counter assistant 59 (35.8) drug for medical abortion in future. The two most significant factors On the job trained 10 (6.0) that independently predicted ever dispensing misoprostol were avail- Total 165 (100.0) ability of, and demand for, the drug. 170  |     Ganle eT al. TABLE  2 Availability and demand for misoprostol by residential area (n=165).a Residential area Availability and demand for misoprostol 1st class 2nd class 3rd class Total Currently stock misoprostol Yes 17 (68) 55 (47.8) 11 (44.0) 83 (50.3) No 8 (32) 60 (52.2) 14 (56.0) 82 (49.7) Demand for misoprostol in facility Daily 2 (8.0) 18 (15.7) 1 (4.0) 21 (12.7) Weekly 4 (16.0) 36 (33.0) 12 (48.0) 54 (32.7) Monthly 9 (36.0) 25 (21.7) 8 (32.0) 42 (25.5) Not sure 10 (40.0) 30 (26.1) 3 (12) 43 (26.1) Never 0 4 (3.5) 1 (4.0) 5 (3.0) Total 25 (100) 115 (100) 25 (100) 165 (100) aAll values are given as number (percentage). The seeming disparity between high demand for misoprostol at medical abortion. Apart from concerns that induced abortion runs community pharmacies and the limited supply of the drug deserves counter to their religious beliefs, many were either unsure of the dose further commentary. Supply of misoprostol from community pharma- to give, or felt it was outside their scope of practice. While it would not cies appears to be limited by two factors: the drug not being stocked be appropriate to expect someone to change their religious beliefs, it at all, and the unwillingness of pharmacy workers to dispense the drug, may be possible to address the other two barriers through reforms and even in situations where it is stocked and there is demand. Availability education. According to Ghana’s abortion law, pharmacists should not of the drug was particularly limited in low- income neighborhoods undertake induced abortion. This may partly explain why many respon- where weekly demands were relatively high. While this disparity could dents felt it was beyond their remit to dispense misoprostol for medical be related to economic factors such as perceived low purchasing abortion. However, given that community pharmacies are increasingly power in low- income neighborhoods which could have led to other serving as important sources of induced abortion services, there is drugs being prioritized over misoprostol because of lack of funds,18 a need to revise Ghana’s abortion legislation to include pharmacists the consequence could be a potential increase in unmet need for safe as providers of induced abortion services. If this legislative reform is abortion services in low- income neighborhoods in the city. This in turn implemented, community pharmacy workers could be trained via both could lead to a rise in the number of unsafe abortions. These findings the regular curriculum and on- the- job training related to appropriate suggest a need for legislative reforms to broaden access to safe legal prescription and dispensing of abortion-i nducing drugs such as miso- induced abortion, or to legalize abortion nationally, ensuring wide- prostol. This could empower community pharmacists to effectively spread provision in public and private facilities in order to bridge the respond to the growing unmet need for safe abortion services. gap between growing demand for induced abortion services on one Regarding respondents’ limited knowledge on the correct dose of hand and limited provision on the other hand. misoprostol to dispense, other than a shortage of pharmacists, this Another important finding relates to the reasons why some com- knowledge gap could be a result of insufficient training in the use of munity pharmacy workers do not want to dispense misoprostol for misoprostol as an abortion-i nducing drug. Insufficient training is more TABLE  3 Misoprostol prescription for medical abortion by qualification of pharmacy employees (n=165).a Medicine counter Variable Pharmacist Pharmacy technician assistant On the job trained Total Ever prescribed misoprostol for induced abortionb Yes 17 (24.6) 10 (37.0) 15 (26.3) 2 (20.0) 44 (26.7) No 51 (74.0) 16 (59.3) 42 (73.7) 8 (80.0) 117 (70.9) Would prescribe misoprostol for induced abortionc Yes 7 (10.1) 6 (22.2) 9 (15.3) 0 (0.0) 22 (13.3) No 61 (88.4) 20 (74.1) 50 (84.8) 9 (90.0) 140 (84.6) Total 69 (100.0) 27 (100.0) 59 (100.0) 10 (100.0) 165 (100.0) aAll values are given as number (percentage). b4 respondents had missing data on whether they had ever prescribed misoprostol for induced abortion. c3 respondents had missing data on whether they would prescribe misoprostol for induced abortion. Ganle eT al.      |  171 TABLE  4 Factors affecting ever prescribing misoprostol by pharmacy workers. Ever prescribed misoprostola,b Variable No (n=117) Yes (n=44) Total (n=161) P value Age, y 0.225 20–24 18 (85.7) 3 (14.3) 21 25–29 43 (78.2) 12 (21.8) 55 30–34 21 (61.8) 13 (38.2) 34 35–39 20 (64.5) 11 (35.5) 31 ≥40 15 (75.0) 5 (25.0) 20 Sex 0.005 Male 48 (64.9) 26 (35.1) 74 Female 69 (79.3) 18 (20.7) 87 Religion 1.000 Christian 108 (72.5) 41 (27.5) 149 Other 9 (75.0) 3 (25.0) 12 Highest level of education 0.269 Tertiary 80 (70.2) 34 (29.8) 114 Secondary 37 (78.7) 10 (21.3) 47 Present qualification 0.577 Pharmacist 51 (75.0) 17 (25) 68 Pharmacy technician 16 (61.5) 10 (38.5) 26 Medicine counter assistant 42 (73.7) 15 (26.3) 57 On the job trained 8 (80.0) 2 (20.0) 10 Stocks misoprostol in facility 0.001 No 69 (87.3) 10 (12.7) 79 Yes 48 (58.5) 34 (41.5) 82 Aware that induced abortion in Ghana is legal in some circumstances 0.203 Yes 64 (68.8) 29 (31.2) 93 No 28 (84.6) 5 (15.2) 33 Don’t know/not sure 25 (71.4) 10 (28.6) 35 Demand for misoprostol 0.001 Daily 10 (47.6) 11 (52.4) 21 Weekly 35 (64.8) 19 (35.2) 54 Monthly 30 (73.2) 11 (26.8) 41 Not sure 37 (92.5) 3 (7.5) 40 Never 5 (100.0) 0 5 Residential Area 0.106 1st class 22 (88.0) 3 (12.0) 25 2nd class 77 (68.1) 36 (31.9) 113 3rd class 18 (78.3) 5 (21.7) 23 a4 respondents had missing data on whether they had ever prescribed misoprostol for induced abortion. bValues are given as number (percentage) unless indicated otherwise. likely given that only 41.8% of the community pharmacies surveyed present study are that Ghana’s Pharmacy Council and Food and Drugs had pharmacists at the time of the study, and the fact that pharma- Authority consider changing the status of misoprostol and related cists are currently not permitted to undertake induced abortions in abortion-i nducing drugs, and providing community pharmacists and Ghana. In order to enhance the knowledge and prescription capabil- related workers with the appropriate education and training to sup- ities of community pharmacy workers, the recommendations of the port women to safely and effectively self- administer the drug when it 172  |     Ganle eT al. TABLE  5 Logistic regression of factors associated with misoprostol dispensing. Ever prescribed misoprostola Variable No (n=117) Yes (n=44) cOR (95% CI) aOR (95% CI) Sexb Male (n=74) 48 (64.9) 26 (35.1) 1 (ref) 1 (ref) Female (n=87) 69 (79.3) 18 (20.7) 0.48 (0.23–0.97) 0.37 (0.16–0.85) Stock misoprostolc No (n=79) 69 (87.3) 10 (12.7) 1 (ref) 1 (ref) Yes (n=82) 48 (58.5) 34 (41.5) 4.89 (2.21–10.83) 4.24 (1.80–10.0) Demand for misoprostold,e Daily (n=21) 10 (47.6) 11 (52.4) 1 (ref) 1 (ref) Weekly (n=54) 35 (64.8) 19 (35.2) 0.46 (0.16–1.30) 0.47 (0.15–1.46) Monthly (n=41) 30 (73.2) 11 (26.8) 0.30 (0.1–1.23) 0.26 (0.08–0.89) Not sure (n=40) 37 (92.5) 3 (7.5) 0.67 (0.14–0.29) 0.51 (0.01–0.25) Abbreviations: aOR, adjusted odds ration; cOR, crude odds ratio; ref, reference category. aValues are given as number (percentage) unless indicated otherwise. bAdjusted for stock misoprostol and demand for misoprostol. cAdjusted for sex and demand for misoprostol. dAdjusted for sex and stock misoprostol. eA response of “never” (n=5) was not included in calculating the odds ratios. is needed. This could contribute to lowering unsafe abortion rates and many low- income settings.14–16 Notwithstanding these limitations, maternal mortality arising from unsafe abortion. important lessons could be learned from the results of the present The present study revealed that demand for misoprostol for med- study to inform policy and practice. ical abortion is high but only half of community pharmacies stock it, and most pharmacy workers do not wish to prescribe the drug. If com- AUTHOR CONTRIBUTIONS munity pharmacies are to effectively contribute to expanding access to safe abortion services in Ghana, the disparity between misoprostol NTB contributed to the conception of the study, the collection and provision and demand needs to be urgently addressed. analysis of the data, and revising the manuscript. JKG contributed to The results of the study should be interpreted in the light of the design of the study, interpretation of the data, and writing and certain limitations. First, the questionnaire used relied on respon- revising the manuscript. EM contributed to the design of the study, dents to recall events that may have taken place several years ago, interpretation of the data, and revising the manuscript. admitting the possibility of recall bias on the part of respondents. Second, given that induced abortion is still a controversial subject CONFLICTS OF INTEREST in Ghana, it could be questioned whether respondents were honest in reporting stocking the drug or dispensing it for medical abortion. The authors have no conflicts of interest. Similarly, responses in relation to future intentions to prescribe the drug as well as correct reporting on demand could be questioned, REFERENCES particularly from respondents who may have indicated that they have never or would not prescribe the drug for medical abortion. 1. Haddad LB, Nour NM. Unsafe abortion: Unnecessary maternal mor- tality. Rev. Obstet. Gynecol.. 2009;2:122–126. A final limitation was that, the study missed an opportunity to learn 2. World Health Organization. Health worker roles in providing safe about provider bias. Provider bias refers to both implicit stereotypes abortion care and post-abortion contraception. [WHO website]. and prejudices about certain personalities, people from certain cul- 2015. http://apps.who.int/iris/bitstream/10665/177628/1/WHO_ tural and socioeconomic backgrounds, as well as understandings RHR_15.11c_eng.pdf?ua=1. Accessed September 30, 2015. 3. Creinin MD. Medical abortion regimens: Historical context and over- of how the world works; healthcare workers could bring provider view. Am J Obstet Gynecol. 2000;183(2 Suppl):S3–S9. bias into service provision as well as into their interaction with cli- 4. Verk J, Zhang J, Olsen J. Medical abortion and the risk of subsequent ents.14,15 These biases may be against a particular method of care adverse pregnancy outcomes. N Engl J Med. 2007;357:648–653. such as induced abortion, a client characteristic or situation, or both, 5. Juarez F, Sundaram A, Ahiadeke C, Bankole A, Singh S, Blades N. The impact of Ghana’s R3M program on the provision of comprehensive and they may not be immediately evident to the providers them- abortion care services. Health Policy Plann. 2015;30:1017–1031. selves.16 This has been shown to be a significant barrier to provid- 6. Wilson KS, Garcia SG, Lara D. Misoprostol Use and Its Impact on ing reproductive health services, including induced abortion care in Measuring Abortion Incidence and Morbidity. In: Singh S, Remez L, Ganle eT al.      |  173 Tartaglione A, eds. Methodologies for Estimating Abortion Incidence and among pharmacies in Bangladesh via mystery client survey. Int J Abortion-Related Morbidity: A Review. New York: Guttmacher Institute; Gynecol Obstet. 2014;124:164–168. 2010:191–204. 16. Lara D, Abuabara K, Grossman D, Díaz-Olavarrieta C. Pharmacy provi- 7. Winikoff B, Sheldon W. Use of medicines changing the face of abor- sion of medical abortifacients in a Latin American city. Contraception. tion. Int Perspect Sex Reprod Health. 2012;38:164–166. 2006;74:394–399. 8. Allen R, O’Brien BM. Uses of misoprostol in obstetrics and gynecol- 17. Smith F. Community pharmacy in Ghana: Enhancing the contribution ogy. Rev. Obstet. Gynecol.. 2009;2:159–168. to primary health care. Health Policy Plann. 2004;19:234–241. 9. Lauro D. Abortion and contraceptive use in Sub- Saharan Africa: How 18. Levandowski B, Voetagbe G, Danquah D, Sneeringer R, Jehu-Appiah women plan their families. Afr J Reprod Health. 2011;15:13–24. K. Misoprostol provision through community pharmacy and licensed 10. Shah IH, Weinberger MB. Expanding access to medical abortion: chemical shops in Ghana. Contraception. 2012;86:300. Perspectives of women and providers in developing countries. Int J 19. Ghana Statistical Service. Ghana Population and Housing Census 2010. Gynecol Obstet. 2012;118(Suppl.1):S1–S3. Accra: Ghana Statistical Service; 2012. 11. Der EM, Moyer C, Gyasi RK, et al. Pregnancy related causes of deaths: 20. Pharmacy Council. Gazette of Pharmacies, 2013. Accra: Pharmacy A 5- year retrospective study. Ghana Med J. 2013;47:158–164. Council; 2013. 12. Aniteye P, Mayhew SH. Shaping legal abortion provision in Ghana: 21. Accra Metropolitan Assembly. City of Accra, Ghana Consulative using policy theory to understand provider-related obstacles to Citizens’ Report Card. 2010. http://www.globalcommunities.org/pub- policy implementation. Health Research Policy and Systems 2013. lications/2010-ghana-citizens-report-card.pdf. Accessed October 14, https://doi.org/10.1186/1478-4505-11-23. Accessed September 2015. 27, 2015. 22. Yamane T. Elementary Sampling Theory. New Jersey: Prentice-Hall; 13. Azhar S, Hassali MA, Izham M, et al. The role of pharmacists in devel- 1967. oping countries: The current scenario in Pakistan. Hum Resour Health. 2009;7:54. 14. Sneeringer RK, Billings DL, Ganatra B, Baird TL. Roles of pharma- SUPPORTING INFORMATION cists in expanding access to safe and effective medical abortion in developing countries: A review of the literature. J Public Health Policy. Additional supporting information may be found online in the 2012;33:218–229. Supporting Information section at the end of the article. 15. Huda FA, Ngo TD, Ahmed A, Alam A, Reichenbach L. Availability and provision of misoprostol and other medicines for menstrual regulation Table S1. Reasons given for decisions surrounding prescribing.