See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/276163027 Causes of Death in Hospitalized HIV Patients in the Early Anti-Retroviral Therapy Era Article · April 2015 DOI: 10.4314/gmj.v49i1.2 CITATIONS READS 3 12 6 authors, including: Ernest Kenu University of Ghana 81 PUBLICATIONS   341 CITATIONS    SEE PROFILE Some of the authors of this publication are also working on these related projects: Audit Project Odense (APO) View project Anthrax in Ghana View project All content following this page was uploaded by Ernest Kenu on 17 October 2017. The user has requested enhancement of the downloaded file. March 2015 Volume 49, Number 1 GHANA MEDICAL JOURNAL CAUSES OF DEATH IN HOSPITALIZED HIV PATIENTS IN THE EAR- LY ANTI-RETROVIRAL THERAPY ERA M. LARTEY1,4, A. ASANTE-QUASHIE2, A. ESSEL3, E. KENU4, V. GANU4 and A. NEEQUAYE1,4 1Department of Medicine, University of Ghana Medical School, College of Health Sciences, Korle Bu, Accra, Ghana 2International Organization for Migration, Accra, Ghana, 3Department of Community Health, Uni- versity of Ghana Medical School, Korle Bu.4Department of Medicine, Korle-Bu Teaching Hospital, Korle-Bu, Accra, Ghana. DOI: http://dx.doi.org/10.4314/gmj.v49i1.2 Corresponding author: Professor Margaret Lartey E-mail address: malart38@yahoo.com   Conflict of Interest: None declared SUMMARY INTRODUCTION Objective: To establish the cause(s) of death HIV/AIDS is among the leading cause of death worldwide, among persons with HIV and AIDS admitted to the with a yearly toll of 3.1million. Most deaths (2.4 million) Fevers Unit of the Korle-Bu Teaching Hospital occur in sub-Saharan Africa.1However, there has been a (KBTH) in 2007 and to determine whether they reduction in AIDS-related deaths and a corresponding in- were AIDS-relatedin the era of availability of crease in non-AIDS-related deaths in the era of highly ac- HAART tive antiretroviral therapy (HAART), especially in devel- Method: Retrospective chart review of all deaths oped nations.2-6 This phenomenon has been attributed to that occurred in the year 2007 among inpatients the increased longevity and better quality of life afforded with HIV infection. Cause of Death (COD) was by HAART. established with post mortem diagnosis, where not available ICD-10 was reviewed independently by As more countries scale up antiretrovirals, the phenomenon two physicians experienced in HIV medicine and a is also being noted across the world in middle and lower consensus reached as to the most likely COD. income countries.3,7,8 AIDS defining illnesses like crypto- Results: In the year under review, 215 (97%) of coccal meningitis, herpes simplex encephalitis, cerebral the 221 adult deaths studied were caused by AIDS toxoplasmosis and extrapulmonary tuberculosis are giving and HIV-associated illnesses. Of these, 123 way to sepsis, cardiovascular diseases, injuries and non (55.7%) were due to an AIDS-defining illness as AIDS defining cancers. described in CDC Category 3 or WHO stage 4. Infections accounted for most of the deaths 158 In developing countries where the majority of HIV- (71.5%), many of them opportunistic 82 (51.8%). infected people live, access to HAART has improved, Tuberculosis was the commonest COD. Clinical however there still remains a substantial unmet need for diagnosis of TB was accurate in 54% of deaths, but antiretroviral therapy. In the treatment 2015 initiative, the was not validated by autopsy in 36% of deaths. UNAIDS recognizes that access to ante-retroviral therapy There were few deaths (14.5%) in patients on (ART) be scaled up so that 15 million persons without HAART. ART have access by 2015. Ninety percent of those without Conclusion: In a developing country like Ghana access who are eligible are found in 30 countries most of where HAART was still not fully accessible, which are low and middle income countries.9 AIDS-related events remained the major causes of death in persons living with HIV. Total scale-up of Initially the high cost and complexity of administration of the ART programme with continuous availability ART in low income countries was considered a barrier but of antiretrovirals is therefore imperative to reduce evaluation of one of the earliest programmes showed simi- deaths from AIDS and HIV associated illnesses. lar virological, immunological and adverse effect outcome There is need for interventions for early diagnosis as in developed countries.10 as well as reduction in late presentation and also better diagnostic tools for tuberculosis. Ghana started its ART programme in 2003 and Korle Bu Teaching Hospital (KBTH) was one of the initial sites to Keywords Cause of death, HIV, AIDS, HAART, offer ART on a large scale. Ghana, Tuberculosis 7 March 2015 M. Lartey et al Mortality in HIV inpatients The study sought to determine the spectrum of was coded according to the International Classification of causes of death (COD) among HIV-infected in- Diseases, 10th Revision (ICD-10). Variables studied in- patients four years after provision of ART at the cluded age, gender, duration of admission at the time of KBTH. death, HAART use, and morbidity. “Morbidity” represent- ed the number of concurrent illnesses present at the time of BACKGROUND death. The Fevers Unit of Korle-Bu Teaching Hospital serves as the national referral centre for HIV- Primary causes of death were classified as “AIDS-related infected patients. Since the first case of HIV was death” (ARD), “HIV-associated death” (HIA), or “Other”. diagnosed in Ghana in 1986, the Unit has provided A death was considered “AIDS-related” when the primary care and support to persons living with HIV/AIDS. COD was an AIDS-defining event as described in Catego- ry 3 of the CDC definition of AIDS 11, or WHO stage 4. In December 2003, the Unit became the third pub- “HIV-associated deaths” included those conditions that lic site to provide HAART in the scale-up of access were HIV related but not CDC category 3 or WHO stage 4. to treatment by the National AIDS Control Pro- A cause of death not directly attributable to either of the gramme (NACP) and its partners. At the time of foregoing was characterized as “Other”, i.e. non-HIV- the study there were about 7000 patients enrolled, associated. This category included conditions such as he- and 3192 of these were receiving HAART. There patic disease or cardiovascular disease from hypertension. were 716 admissions to the ward in the year under study. Causes of death were also sub-categorized by organ sys- tems and stratified into “Opportunistic infection”, “Oppor- We explore here in detail the causes of death tunistic malignancy”, or “Other”. Examples of conditions among HIV-infected patients admitted at the Korle- falling into this last category were malignancies such as Bu Teaching Hospital in the year 2007, 4 years hepatocellular carcinoma and metastatic choriocarcinoma. after the national HAART programme was launched at the Fevers Unit. Data management Data was entered into an excel workbook and cleaned. METHODS Missing data were filled in where available and duplica-tions removed. The data was then exported into STATA Study Design version 8.2 and variables coded for analysis. This was a retrospective study in which the medical and mortality records of all HIV-infected patients Statistical Analysis who died at the Fevers Unit during the study period Statistical comparisons were made using the chi-squared (January 2007 to December 2007) were reviewed. test of hypothesis or Fisher’s exact test, where appropriate. Data from multiple sources were used and triangu- P values were two-tailed, and values of <0.05 were consid- lated. The multiple sources included one or more of ered statistically significant. All analyses were done using the following: death certificates, medical charts, a standard statistical package, STATA, version 8.2 (Stata and autopsy reports, where available. The multiple Corp., College Station, Texas). sources also included electronic data capture from the database as well as manual records from the Ethics patient folders which included more details of out- The study protocol was reviewed and approved by the patient as well as all details of inpatient admission. Ethical and Protocol Review Committee of the University Some of the inpatient records were doctors’ notes, of Ghana Medical School. nurse’s notes and results of laboratory tests. Causes of death (COD) were collated from the death certif- icates and matched with autopsy results. RESULTS Demographic Characteristics Where autopsy results were not available, two phy- Between January and December 2007, a total of 716 HIV sicians experienced in HIV medicine using the and AIDS patients were admitted to the Fevers Unit of medical records of admission and concordance which 221 of these patients died on admission giving a established did independent determination of the crude mortality rate of 31.5%. Autopsy records were avail- most likely cause of death. In cases of discordance, able for 135 (61%) of the deaths. The mean age was 39.7 a third experienced physician reviewed the records (SD 9.01) and the age range was 18-80 years. Patients in and consensus was reached as to the most likely the reproductive age group (15-49) made up 187(84.6%) primary and secondary CODs, taking into account whilst females made up 115 (52%) of the population. the morbidity at the time of death. Cause of death 8 March 2015 Volume 49, Number 1 GHANA MEDICAL JOURNAL The majority of patients 116 (53%) spent between Infections were again the commonest cause of HIV associ- one and six days on admission before death. Only ated deaths majority of which were lobar or bronchopneu- 32 (14.5%) of the patients were on HAART at the monia. This was followed closely by anaemia. time of death (Table 1). Accuracy of Pre-mortem Diagnosis Causes of Death Autopsy findings were fully in agreement with pre-mortem Out of the total deaths, 215 (97.3%) were caused diagnoses in 24.44% (33) of cases, partial agreement in by conditions associated with HIV and AIDS and 51.85% (70) and not in agreement with 23.70% (32) of 6(2.7%) were caused by non-HIV related condi- cases. tions. Of the HIV and AIDS related deaths, 123 (57.21%) were due to an AIDS-defining illness as Table 2 AIDS Related Deaths described in Category 3 of the CDC definition of Infections N=110 N (%) AIDS and therefore were categorized as AIDS- Disseminated TB 45(36.5) related deaths (ARD) (Table 2) whereas 92 deaths TB meningitis 7 (6.2) (42.79%) were non AIDS related and classified as Pulmonary TB 17 (15) HIV-associated deaths (HIA). Cerebral Toxoplasmosis 26 (23) Cryptococcal meningitis 4 (3.5) Bacterial meningitis 3 (2.6) Table 1 Comparison of patients dying from AIDS Chronic diarrhea 8 (7) Related deaths (ADR) and HIV Associated deaths Malignancies N-13 (HIA) Kaposi’s sarcoma 10 (8.6) Factor(at time AIDS re- HIV associ- p- value Lymphoma 3 (2.6) of death) lated ated deaths deaths No No(%) (%) (N=92) Table 3 HIV Associated Deaths (N=123) HIV Associated deaths N-92 N (%) Age (yrs) Infections 44 (47.8) 15-49 103 (83.7) 79 (85.9) 0.81 Anaemia 25 (27.2) >49 20 (16.3) 13 (14.1) Miscellanous 19 (20.7) Sex Malignancy 4 (4.3) Males 61 (49.6) 42 (44.7) 0.71 Females 63 (50.4) 50 (54.3) Duration of Admission at time of death Accuracy of TB Diagnoses <24 hrs 2 (2) 6 (6) 0.85 The ante mortem diagnosis of 69 cases of TB was validat- 1-6 days 65 (55.5) 50 (49.5) ed using post mortem diagnosis. Clinical diagnosis of TB 7 days or more 50 (42.5) 45 (44.5) was accurate in 37(54%) of TB deaths, but was not validat- HAART use Yes 16 (13.6) 16 (15.8) 0.80 ed by autopsy in 25 (36%) of cases, revealing a considera- No 101 (86.4) 85 (84.2) bly high index of suspicion for the disease. The diagnosis Morbidity was missed in 7(10%) of patients. 0 concurrent 48 (41.1) 27 (26.7) <0.01* illnesses 1 concurrent 45 (38) 42 (41.6) DISCUSSION illness The causes of death for HIV-infected patients admitted to 2 concurrent 23 (20) 20 (19.8) the Fevers Unit in 2007 were examined. The data showed illnesses that AIDS-related events continued to be the major causes 3 or more con- 1 (0.9) 12 (11.9) of death, in contrast to recent studies conducted in industri- current illnesses *statistically significant at p<0.05 alized countries which had noted a shift in the causes of death toward non-HIV-related causes since the introduc- Tuberculosis was the commonest cause of death tion of HAART. 2,12-14 accounting for 69 (57.7%) of all deaths. This was 12 followed closely by cerebral toxoplasmosis and Sackoff noted in his study that from 1999-2004, the per- Kaposi’s sarcoma. centage of deaths from non-HIV causes had increased by 32%, whereas Palella14 also found that the proportion of HIV associated deaths consisted mainly of infec- deaths attributable to non AIDS diseases had increased and tions (47.8%), anaemia (27.2%), malignancy were predominantly hepatic, cardiovascular, pulmonary (4.3%) and a large miscellaneous group (20.7%) and non AIDS defining cancers. In the study by De Ollala 4 consisting of venous thromboembolism, pulmonary et al , cause of death from non HIV related causes was oedema and others (see Table 3). 28% with cancers contributing 20% and liver diseases 18.8%. 9 March 2015 M. Lartey et al Mortality in HIV inpatients In this cohort however, non-AIDS-related causes of With regard to demographic characteristics, patients dying death such as non-opportunistic malignancies, car- from ADIs differed from those dying from non-ADIs only diovascular, hepatic and renal diseases were un- in morbidity. This finding was rather surprising as we ex- common. On the other hand, infections were the pected the opposite, given our definition of morbidity as leading causes of death in this population, notably number of concurrent illnesses at the time of death. In a opportunistic infections. Tuberculosis was the sin- study on AIDS and non-AIDS mortality in the era of an- gle most important opportunistic infection causing tiretroviral therapy by Falster et al 16, they found that of death. 215 deaths, 89 were AIDS related, 97 non-AIDS related and 29 were unknown. They found age greater than 50 Disseminated TB was the commonest form, with years and CD4 counts > 100 increased the risk of non pulmonary TB running a close second and TB AIDS deaths (HR 4.99) whereas CD4 ≤100 and viral load meningitis, third. Likewise in a study conducted in >10,000 increased the risk of AIDS deaths (HR 4.21). Burkina Faso on determinants and causes of mor- Most of our patients did not have CD4 counts and viral tality in HIV patients on HAART between 2003 load tests were not available at the time. and 2008, a multivariate analysis showed that clini- cal stage, BMI, CD4 count and treatment regimen Most of the patients in this study were not on HAART at were significantly associated with death. Common the time of their death, suggesting a low admission rate of causes of mortality were wasting syndrome, tuber- patients on HAART. Even though this rendered analyses culosis and anemia.7 by HAART use inconclusive, it shows the benefit of HAART in reducing morbidity severe enough to warrant The pattern of diseases causing death in Ghana and admission and subsequently mortality. Increased use of Burkina Faso could be attributed to the low pene- HAART is likely to reduce mortality as shown in other tration of antiretroviral therapy amongst the popu- studies around the world.2,12-14 lation studied and also the fact that these countries still continue to have a high burden of communica- Some of the admissions of patients on HAART were at- ble diseases as compared to non-communicable tributed to HAART related anaemia caused by zidovudine, diseases. The pattern also depicts the severe im- one of the first line drugs. Availability and accessibility of mune suppression that exists in the in –patients. blood and blood products continue to be a challenge for patient care. The relatively short duration of admission for In a study by Agaba et al 15 on predictors of mortal- most patients reflects the late presentation of HIV-infected ity in hospitalized HIV patients in Nigeria, tubercu- patients at the health facility hence the high mortality rate. losis was the common diagnosis accounting for 33.6% of the deaths. Pulmonary disease accounted Limitations of this study include the frequent inadequacy for 53% and extra-pulmonary 47% of TB cases. In of death certificates and medical notes to identify causes of this study, tuberculosis was again the commonest death; these are imperfect methods. The autopsy rate was cause of death accounting for 57.7% of all deaths. just over 60% in this study. Another limitation was that However in contrast to findings from Agaba, dis- immunological and virological parameters were not readily seminated and extrapulmonary accounted for 75% available for analysis due to late presentation, short dura- of cases whilst pulmonary TB was 25%. tion of stay and financial challenges. Virological tests were not available at the time. Also, being a tertiary referral cen- In this study, we have shown that a significant pro- tre, mortality at the Fevers Unit may not be representative portion of TB deaths (10%) were not diagnosed of HIV mortality in the community but rather in hospitals. prior to death. This highlights the need for better Mortality in the community can only be ascertained using diagnostic tools for TB a situation which to date verbal autopsies. has only marginally improved. In about three- quarters of the cases studied, at least one pre- CONCLUSION mortem finding was confirmed by autopsy, attest- The overall mortality rate for 2007 was 31.5%. AIDS- ing to the high clinical acumen of the physicians related causes accounted for 55.7% whereas HIV- involved and the specificity of the post mortem associated causes accounted for 41.6% of total deaths. To- diagnosis. This is in contrast to the study on causes gether AIDS and HIV related deaths formed 97% of the of mortality conducted by Sackoff, where he ob- deaths showing that the transition from HIV related deaths served that death certificates issued lacked specific- to causes of deaths similar to those of the general popula- ity of the cause of death.12 tion had not yet occurred in the Unit. Autopsy findings confirmed at least one pre-mortem diagnosis in 76% of deaths. Tuberculosis was the single most important cause of death in the study population. 10 March 2015 Volume 49, Number 1 GHANA MEDICAL JOURNAL Accuracy of TB diagnosis was average and still mortality and changing patterns of causes of death in needed to be supported by better diagnostic tools. the Swiss HIV Cohort Study: Causes of death and HIV In a developing country like Ghana where HAART infection. HIV Med. 2013 Apr;14(4):195–207. is yet to be widely accessible to PLHIVs, HIV- 7. Kouanda S, Meda IB, Nikiema L, Tiendrebeogo S, related deaths remain the major cause of death. Doulougou B, Kaboré I, et al. Determinants and caus- es of mortality in HIV-infected patients receiving an- RECOMMENDATIONS tiretroviral therapy in Burkina Faso: a five-year retro- We recommend further scale-up of the ART pro- spective cohort study. AIDS Care. 2011 Dec 7;1–13. gramme, early diagnosis of individuals and con- 8. Yang C-H, Huang Y-F, Hsiao C-F, Yeh Y-L, Liou H- sistent supply of drugs and other consumables to R, Hung C-C, et al. Trends of mortality and causes of reduce HIV related mortality, improve quality of death among HIV-infected patients in Taiwan, 1984- life and increase life expectancy. Better diagnostic 2005. HIV Med. 2008 Aug;9(7):535–43. tools for TB are needed to assist clinicians and pa- 9. JC2484_treatment-2015_en.pdf [Internet]. [cited 2013 tients. Aug 27]. Available from: http://www.unaids.org/en/media/unaids/contentassets/ ACKNOWLEDGEMENTS documents/unaidspublication/2013/JC2484_treatment- The authors wish to acknowledge all HIV-infected 2015_en.pdf patients and the staff of Fevers Unit particularly 10. Djomand G, Roels T, Ellerbrock T, Hanson D, Dio- Perfect Dzandu for collating and cleaning the data. mande F, Monga B, et al. Virologic and immunologic outcomes and programmatic challenges of an an- REFERENCES tiretroviral treatment pilot project in Abidjan, Côte d’Ivoire. AIDS Lond Engl. 2003 Jul;17 Suppl 3:S5–15. 1. Etard J-F, Ndiaye I, Thierry-Mieg M, Guèye 11. Schneider E, Whitmore S, Glynn KM, Dominguez K, NFN, Guèye PM, Lanièce I, et al. Mortality Mitsch A, McKenna MT, et al. Revised surveillance and causes of death in adults receiving highly case definitions for HIV infection among adults, ado- active antiretroviral therapy in Senegal: a 7- lescents, and children aged <18 months and for HIV year cohort study. AIDS Lond Engl. 2006 May infection and AIDS among children aged 18 months to 12;20(8):1181–9. <13 years--United States, 2008. MMWR 2008 Dec 2. Crum NF, Riffenburgh RH, Wegner S, Agan 5;57(RR-10):1–12. BK, Tasker SA, Spooner KM, et al. Compari- 12. Sackoff JE, Hanna DB, Pfeiffer MR, Torian LV. sons of causes of death and mortality rates Causes of death among persons with AIDS in the era among HIV-infected persons: analysis of the of highly active antiretroviral therapy: New York City. pre-, early, and late HAART (highly active an- Ann Intern Med. 2006 Sep 19;145(6):397–406. tiretroviral therapy) eras. J Acquir Immune 13. Bonnet F, Morlat P, Chêne G, Mercié P, Neau D, Defic Syndr 1999. 2006 Feb 1;41(2):194–200. Chossat I, et al. Causes of death among HIV-infected 3. Sun Hee Lee, Seung Geun Lee, Dong Hwan patients in the era of highly active antiretroviral thera- Chen, Dong Sik Jung, Chi Sook Moon, Ji py, Bordeaux, France, 1998–1999. HIV Med. Young Park, Joo Seop Chung, Ihm Soo Kwak, 2002;3(3):195–9. and Goon Jae Cho, Kye-Hyung Kim,. Trends 14. Palella FJ Jr, Baker RK, Moorman AC, Chmiel JS, of Mortality and Cause of Death among HIV- Wood KC, Brooks JT, et al. Mortality in the highly ac- Infected Patients in Korea, 1990-2011. J Kore- tive antiretroviral therapy era: changing causes of an Med Sci 2013 28 67-73. 2013;28:67–73. death and disease in the HIV outpatient study. J Ac- 4. De Olalla P, Novoa A, Clos R, Orcau A, Ro- quir Immune Defic Syndr 1999. 2006 Sep;43(1):27– driguez-Sanz M, Cayla J. Increase in the Non- 34. HIV-Related Deaths Among Aids Cases in the 15. Agaba AP, Digin E, Makai R, Apena L, Agbaji O, HAART Era. Curr HIV Res. 2008 Jan Idoko J, Murphy R, Kanki P. Clinical characteristics 1;6(1):77–81. and predictors of mortality in hospitalised HIV- 5. Simmons R, Ciancio B, Kall M, Rice B, infected Nigerians. J Infect Dev Ctries 2011; 5(5):377- Delpech V. Ten-year mortality trends among 382. persons diagnosed with HIV infection in Eng- 16. Falster K, Choi JY, Donovan B, Duncombe C, Mulhall land and Wales in the era of antiretroviral ther- B, Sowden D, Zhou J, Law MG. AIDS and non-AIDS apy: AIDS remains a silent killer: Mortality in related mortality in the Asia Pacific region in the era the era of ART. HIV Med. 2013 Jun; 14: 596– of combination antiretroviral treatment. AIDS. 2009 604 November 13; 23(17): 2323-2336 ✪ 6. Weber R, Ruppik M, Rickenbach M, Spoerri A, Furrer H, Battegay M, et al. Decreasing 11 View publication stats