Identifying Factors Related to the Survival of AIDS Patients under the Follow-up of Antiretroviral Therapy (ART): The Case of South Wollo

Transcription

1 International Journal of Data Envelopment Analysis an *Operations Research*, 014, Vol. 1, No., 1-7 Available online at Science an Eucation Publishing DOI: /ijeaor-1-- Ientifying Factors Relate to the Survival of AIDS Patients uner the Follow-up of Antiretroviral Therapy (ART): The Case of South Wollo Jemal Ayalew 1, Helen Moges 1, Omprakash sahu,*, Anteneh Worku 1 Department of Statistics, College of Natural Sciences, Wollo University, Ethiopia Department of Chemical Engineering, KIOT Wollo University Kombolcha Ethiopia *Corresponing author: ops011@gmail.com Receive April 01, 014; Revise March 09, 014; Accepte July 17, 014 Abstract The expansion of HIV/ADIS epiemic has now become a burning issue globally. HIV infection has change from a fatal conition to a manageable chronic illness mainly ue to the evelopment of antiretroviral therapy (ART). Even if ART treatment has shown significant clinical importance by meeting the goal of therapy, we are still facing a number of eaths ue to certain socio-economic, emographic, behavioral risk an health factors. The aim of this stuy is for ientifying eterminant factors for the Survival of HIV/ADIS Infecte Patients uner the follow-up of antiretroviral therapy (ART) at Boru Mea an Dessie Referral Hospitals an Kombolcha Health Center. The ata for this research were collecte uring the follow-up time from January 1, 008 to December 31, 011. Out of a population of HIV-patients who were taking antiretroviral therapy in the hospitals an health center in that perio, 654 patients were selecte base on simple ranom sampling technique for this stuy. The stuy subjects were people in the age range from 15 to 75 years. The Kaplan-Meier Metho was employe to estimate survival; the Cox Proportional Hazars Regression Metho was use to ientify eterminants of survival. After initiation of the antiretroviral treatment, HIV-positive patients the estimate meian survival age was foun to be 48 months (CI: years). Living rural, orer baseline age, not working ue to illness, smaller CD4 count, weight, an lymphocyte count, HIV-TB co-infection eveloping an being in WHO clinical Stage IV were ientifie as a ocumente risk factors for shortene the survival experience of the HIV-patients who are in care of ART. The mortality among our patients was comparable to that reporte from other low-income countries. Earlier initiation of ART may reuce the high mortality rates observe. Similar stuies in the future nee to consier preictors in aition to those consiere in this stuy an the clinical avantage of ART that brings improvement on the clinical characteristics in the follow up perio. Keywors: antiretroviral, bacterial, infection, global, patients Cite This Article: Jemal Ayalew, Helen Moges, Omprakash sahu, an Anteneh Worku, Ientifying Factors Relate to the Survival of AIDS Patients uner the Follow-up of Antiretroviral Therapy (ART): The Case of South Wollo. International Journal of Data Envelopment Analysis an *Operations Research*, vol. 1, no. (014): 1-7. oi: /ijeaor Introuction Acquire Immune Deficiency Synrome (AIDS) which is believe to be cause by the Human Immunoeficiency Virus (HIV) has been the major problem worlwie. The rate of sprea of the HIV/AIDS epiemic has reache a shocking level. The expansion of the epiemic has now become a burning issue globally an this is particularly so more in eveloping countries; specially, in Sub Saharan Africa [1]. In fact, Sub-Saharan Africa accounts for.4 million infections, which is about 67% of the total HIV buren. The number of people estimate to acquire new infections is aroun 1.9 million accounting for 68 % of the total number of new infections []. This was note as a significant reuction in the number of new cases since 001. However, it was also reporte that HIV/AIDS has become the leaing cause of eath in the region [1]. Ethiopia is one of the Sub-Saharan African countries most severely affecte by the HIV/AIDS panemic. Currently, the national ault prevalence rate is estimate at.3 percent an an estimate number of 1. million people are living with HIV/AIDS. An estimate 67,000 lost their lives ue to AIDS at the en of 007 [3]. The worl is now above the thir ecae of AIDS epiemic since elapse. During these years HIV infection has change from a fatal conition to a manageable chronic illness mainly ue to the evelopment of antiretroviral therapy (ART). ART for the treatment of HIV infection has been shown to profounly alter HIV isease progression, incluing incience of opportunistic infections in people living with HIV [4]. The goal of this therapy is to improve survival; to reuce HIV associate morbiity an mortality, to increase the quality of life, to

2 International Journal of Data Envelopment Analysis an *Operations Research* restore immune function an to achieve maximal an sustaine suppression of viral replication [5]. By 010, WHO has planne to put 9.8 million people on ART with the goal of proviing universal access to HIV care an ART [6]. Accurate estimates of expecte survival times, survival rates of AIDS patients an variables that influence survival are important for projecting future numbers of AIDS cases, increasing unerstaning of the pathophysiology of the isease, clinical ecision making an planning health service interventions [3,7]. In aition for monitoring the progress of the HIV/AIDS epiemic as new treatments, previous stuies have shown that survival stuies of patients with AIDS provie information on the response of ifferent subgroups to the synrome, an enable mathematical moelers an health care professionals to estimate future nees [8]. However, most of the stuies in Ethiopia focuse on the prevention an factors that increase the chance of contracting the isease. It can be sai that limite works were conucte on the survival of HIV positives taking ART [9]. Although the current HIV/AIDS surveillance estimates inicate some encouraging signs in that the epiemic is stabilizing, the observe changes are not sufficient enough compare to the esire goals of the response against the epiemic. It is believe that, in resource poor countries like Ethiopia the survival of patients with AIDS treate with ART epens on a variety of factors, which may also vary greatly with economic, emographic, behavioral risk an health factors. In other wors, even if ART treatment has shown significant clinical importance by meeting the goal of therapy, we are still facing a number of eaths that can otherwise be avoie by appropriate interventions on certain socioeconomic, emographic, behavioral risk an health factors. Thus this stuy was unertaken with objectives to ientify preictors that have impact on the survival of HIV AIDS patients with the hope that the results woul contribute to existing knowlege.. Metho.1. Stuy Area The stuy was conucte in selecte governmental health institutions in South Wollo, Amhara region, Ethiopia from January 01 to December 31, 01. It is 401 Kilo meters far from Ais Ababa the capital city of Ethiopia. The selecte institutions are Dessie referral Hospital, Boru Mea Hospital an Kombolcha Health Center... Data The target populations for this stuy were patients uner the follow up of ART at Boru Mea an Dessie Referral Hospitals an Kombolcha Health Center from 1 January, 008 to 31 December, 011. At the Hospital s ART clinic the ata were recore using the stanarize ata collection formats an registers prepare by the Ministry of Health. Data recoring was one by health officers an nurses working in the clinic. The examining meical octors also recore follow up information about their patients. Base on this recor of the patients who ha complete information, the variables which are important for the stuy was collecte using the patients meical registration number (MIR) without any irect contact with the patients, instea by communicating with the nurses an counselors to get the meical recor an other information important for the stuy..3. Sample Size In etermining require sample size there are a number of issues/points one has to take into account. Some of the issues are: objective of the research, esign of the research, cost constraint, egree of precision require for generalization an etc. Accoringly, the sample size etermination formula Z p(1 p) n = 1 Z p(1 p) 1+ 1 N (1) Above equation was aopte for this stuy (Cochran, 1977), where Z is the upper α/ points of stanar normal istribution with α =0.05 significance level, which is Z =1.96. The egree of precision is taken to be The parameter p represents proportion of eath. P = 0.10 is use in this stuy obtaine from previous stuy at Tikur Anbessa Specialize Hospital, Ais Ababa [9]. Accoringly, the sample size using the given formula becomes n = % of the sample size, which is 60, is ae to the etermine sample size 594 to compensate for the ranom errors an the sample size, with population size N=7165 for the current stuy become 654. Next, we carrie out sample size allocation to each Hospital an health center with proportional allocation. Then, patients were selecte from each Hospital an Health Center using systematic sampling base on their Meical Registration Number (MRN) which is given to each patient who is taking ART..4. Variables of the Stuy The response variable in this research that was the survival time was efine as the number of months from the month of enrollment of a patient in the HIV-care till one of the events eath, lost to follow up, roppe out, stoppe, transferre out to other health centers or hospitals occurre. This meant that the survival ata stuie here were right-censore. The preictor variables relate to the social, emographic, meical an clinical backgroun of the patients having these respective classifications; gener (male, female), resience (rural, urban), age at the start of ART (in full year), marital status at the start of treatment (single, separate, ivorce, wiowe), level of eucation at the start of treatment ( no eucation, primary, seconary an above), CD4 count at the start of treatment(/mm3), weight at the start of ART (in kilograms), employment status at the start ART (unemploye, not working ue to illness, on the job), total lymphocyte count at start the treatment, WHO clinical stage at start the treatment (stage I, stage II, stage III, an stage IV), an TB status at starting ART (negative, positive).

3 3 International Journal of Data Envelopment Analysis an *Operations Research*.5. Statistical Analysis The statistical analytic metho use in this stuy is known as Survival Analysis. Survival ata analysis involves the moeling an analyses of ata that have a principal en point the time until an event occurs (time-toevent ata). Survival Analysis consiers conitional information on the remaining time of a subject s survival given current survival time. Survival ata were censore in the sense that they i not provie complete information since, for a variety of reasons, subjects of the stuy may not have experience the event of interest. The existence of variables that change over time is also a istinguishing feature in survival analysis. Descriptive analysis of survival ata utilizes nonparametric methos to compare the survival functions of two or more groups. The Kaplan-Meier estimator (prouct-limit-estimator) of the survival function [10] was employe for this purpose. The log-rank test was utilize to test whether observe ifferences in survival experience between/among the groups was significant or not. The multivariable moel use was the semi-parametric regression moel known as the proportional hazars regression (PHR) moel [11]. When a stuy involves multiple characteristics, appropriate statistical techniques must be use to select variables that have significant effects on survival an which are juge to be clinically meaningful for inclusion in a PHR moel. Hence, the moel evelopment process ientifies the relevant variables following moel scrutiny as iscusse in [1]. 3. Results 3.1. Results of the Descriptive Statistics A sample of 654 of the 7163 patients was selecte that were followe uring 1 January, 008 to 31 December, 011 using an appropriate sample size etermination formula. Of those, about 87% were right-censore (ropout, transferre, loss an alive till the stuy perio) an the remaining 13% uncensore (Die). A summary of the ata for each level of variables is provie in Table 1. In orer to investigate if there is significant ifference between the survivals of a patient between categories of covariates, Kaplan-Meier survivor estimates all significant covariates in the log-rank test. In line with this, the logrank statistical test is mae in Table 1. Table gives the mean survival time an the results base on the log-rank test for each covariate. From the mean survival time in Table HIV-positive patients live for an average of months (CI: months. The p-values in Table show ifferences in survival experience between two or more levels of preictors. All preictors with the exception of gener an eucation level manifest ifferences in levels of survival functions. A résumé emanating from escriptive analysis with reference to the seven preictors (leaving out sex, stuy site an eucation level) that manifest ifferences in survival are provie as follows. Marrie patients live for an average of six months longer than wiowe patients whereas they live for an average of four months longer than single an ivorce patients. Patients who were from rural resience, live on average three months shorter than those from urban resience. Patients who are on job an unemploye live on average 18 an 17 months longer than those in t work ue to illness, respectively. Table 1. Socio-emographic an clinical characteristics of HIVpositive patients at ART start in Boru Mea Hospital, Dessie Referral Hospital an Kombolcha Health Center, , South Wollo, Amhara Region (n =654) Variable Category Censore Faile Tota l Percent censore Gener Male Female Boru Mea Hospital Dessie Stuy area Referral Hospital Kombolcha Health Center Age Group >= Resience Rural Urban Single Marital Marrie Status Divorce Wiowe Illiterate Level of Primary Eucation Seconary an above Base line < 00/mm CD4 Count >= 00/mm < Base Line weight >= Unemploye Employmen t Status Not working ue to illness On job Stage I WHO Stage II Clinical Stage III Stage Stage IV TB Status Negative Positive < 100 Base line cells/mm lymphocyte >= 100 count cells/mm HIV Patients, who evelope TB, ha shorter survival time than not evelope TB: 1 months shorter than patients in t evelop TB. HIV-positive patients, with WHO clinical stage IV ha an average of 8, 6 an 5 months shorter survival time than those in stages II, III an I, respectively. Those patients in clinical stage II ha an average of 3 an months longer survival time than stage I an III, respectively. Patients were having weights 45 an above kilograms live four months longer on average than those having weight below 45 kilograms. Patients, who ha ages 15 to 9 years an 30 to 40 years, respectively, live 3 an 5 months longer on average compare with those people who ha ages greater than 40 years. Those patients having base line CD4 cell count above 00/mm 3 live three months longer on average than patients with base line CD4 cell count below 00/mm 3. Similarly patients having base line lymphocyte count greater or equal to 000 cells/mm 3 ha on average 8 months longer than those ha less than 000 cells/mm 3 base line lymphocyte count.

4 International Journal of Data Envelopment Analysis an *Operations Research* 4 Table. Mean survival time an Log-rank test p-values base on socio-emographic an clinical characteristics ata of HIV-positive patients at ART initiation in Selecte South Wollo Hospitals an health Centers, Amhara Region, , (n =654) Variables Mean Survival time Chi-square Log-rank p-value Sex Female Male Age >= Marital Status Single 40.3 Marrie Divorce Winowe Eucational level No eucation Primary Seconary an above Resience Rural Urban Employment Status Unemploye Not working ue to illness On Job TB Status Negative Positive Baseline CD4 Count < >= Baseline Weight <= > WHO Clinical Stage Stage I Stage II Stage III Stage IV Baseline Lymphocyte count < >= Table 3. Estimate parameters for the preliminary Proportional Hazars Regression Moel Containing variable significant at 5% level in the bivariate analysis for the ata from selecte South Wollo Health centers an Hospitals, Amhara region, , (n =654) Variable B SE Wal f Sig. HR 95.0% CI for HR Age_cat Age_cat(1) Age_cat() Resience Marital_Status Marital_Status(1) Marital_Status() Marital_Status(3) CD weight employee employee(1) employee() Lymphocyte count WHO WHO(1) WHO() WHO(3) TB Status Results of the PHR Moel The Cox moel proceure that inclues moel selection, tests, iagnosis an fit confirme that there were no problems with regar to interactions of main effects an confouning (see appenix B). Therefore, the results in Table 3 an Table 4 are base on the main effects an the following elaboration etails survival experience base on estimate hazar ratios (HR). It shoul be pointe out that variables with p- values below 0.05 were consiere as statistically significant. The relationship between each covariates an survival time of AIDS patients which are significant using a moest level of significance 5% to be inclue for further investigation in the multiple covariates moel are presente in Table 3. P-value in Table 3 inicates survival of the patients is significantly relate with all the propose covariates except gener, stuy area an eucation level of the patient.

5 5 International Journal of Data Envelopment Analysis an *Operations Research* Table 4. Estimate parameters for the final Proportional Hazars Regression Moel for the ata from selecte South Wollo Health Centers an Hospitals, Amhara Region, , (n=654) Variable B SE Wal f Sig. HR 95.0% CI for HR Age group >= Resience Rural Urban 0 1 CD4 Count weight Employee- Status Unemploye Not working ue to-illness On Job 0 1 Lymphocyte Count WHO clinical stage Stage I Stage II Stage III Stage IV 0 1 TB Status Negative Positive 0 1 The covariates in Table.3 were further investigate in the final moel an seven were statistically significant at 5 % level of significant. Details of the variables of the final moel were isplaye in Table.4. Interpretations were mae using final moel as follows Socio-Demographic Characteristics The referent category for age was the age group greater than 40 years. The estimate HRs for the age groups (15-9) an (30-40] were 0.60 (CI: 0.356, 1.01) an (CI:. 83,. 846), respectively. The age group (30-40] was significant (p-value = 0.01), meaning that patients whose age greater than 41 years were ying at a rate of 48.9% greater than age group (30-40). The hazar rate of rural resience patients was 1.74 times greater than urban resience patients (CI: 1.109,.74). For not working ue to illness patients, the estimate HR was 6.19 (CI: 1.96, 6.68) as compare to the referent category on Job at the time of starting ART. The status of being unemploye i not have a significant contribution at 0.05 level. A 5 kilogram increase in weight resulte in an estimate HR of (CI: ) Clinical Characteristics less in CD4 count (HR: 0.998; CI: 0.996,. 999; p=0.005) an lymphocyte count (HR:. 969; CI:. 945,. 994; p=0.015) at start of ART were significant risk factors associate with the survival of HIV patients. The referent category pertaining to the WHO clinical stage was stage IV. Stage II an III are statistically significant an estimate HRs for those stage II an III, respectively, are 0.4 (CI: ) an 0.4 (CI: ). The estimate hazar ratio for Negative TB status was (CI: ) relative to the category Positive TB status, meaning that patients TB evelop were ying at a rate of 53.4% greater than not evelop TB. 4. Discussion This 4-year retrospective cohort stuy of HIV/AIDS patients on ART gives an insight into survival an its eterminants in a hospital setting in Ethiopia. In this cohort, 14.1% of the patients ie even if they were taking ART. The overall mortality rate was higher compare to other stuies in Ethiopia, in which it was shown that in Ais Ababa Tikur Anbessa Specialize Hospital an at two istrict hospitals, Assela an Shashemene, locate in the southern part of the Oromiyaa region, 10.0 an 10.3%, respectively [9,13]. On the other han it was smaller compare to other stuies conucte outsie Ethiopia in which in Korea, Tanzania an Cameroon, 0.8, 9.7 an 3%, respectively [14,15,16]. Similarly, base on 564 sample HIV patients who are taking ART, the mean an meian survival time was almost 4 an 48 months. The meian survival time was similar with [9] a four year perio stuy but larger than [16] a five year follow up stuy whereas the mean survival time in the current stuy was smaller than [9] an larger than [4] a three year follow up retrospective cohort stuy. Rural HIV-positive patients ha shorter survival on average compare to urban. It was observe that the survival time of patients uner ART varie along ifferences in employment status. Not working ue to illness patients ha the least survival time compare to patients who are on job. Those patients aging in between 30 an 40 years ha on average better survival experience than patients aging greater than 40 years. People living with HIV/AIDS uner ART i not evelop TB ha better survival time on average than patients who were evelop TB. In this stuy baseline CD4 count, baseline weight an baseline lymphocyte count were etermine as risk factors for the survival of HIV patients. Thus, as the hazar rate of eath risk are high for those with lower baseline weight, CD4 an lymphocyte count, the lower the anger of being at risk of HIV eath. Accoringly, baseline CD4 cells counts showe a strong influence on the survival status; patients with counts of larger/mm 3 ha higher survival experience than those with counts smaller/mm 3. Baseline Weight showe a strong influence on the survival status; a patient with 5 kg weight gain reuces mortality by 16.5 percent. Similarly, baseline lymphocyte count showe a strong influence on the survival status; patients with counts of higher cells/mm 3 ha higher survival experience

6 International Journal of Data Envelopment Analysis an *Operations Research* 6 than those with counts smaller cells/mm 3. Patients who were in WHO clinical Stage IV ha the highest risk of mortality than stage II an III. Resience has a significance influence on survival, meaning that HIV patients living urban or near town ha lower risk of mortality compare to far from town or rural [18,19,0]. The current stuy also shows that patients living rural experience higher mortality rate than urban. Results in [9,17,0] showe that the survival time was reuce among iniviuals infecte at oler ages. This stuy has also come up with similar conclusions as in [9,17,0]. The stuies [4,9,16,18,19,1] showe that CD4 count is a laboratory preictor of mortality in the sense that higher CD4 counts are associate with longer survival time. The current stuy concurs with conclusions above. Weight loss is a cause for reuce longevity [9]. This stuy agrees with the foregoing conclusion; it also shows that a 5 kg weight gain reuces mortality by 16.5% percent. The finings of the stuy [4,17,19] showe that WHO clinical stage remaine significantly associate with survival of patients taking ART. It was evient that patients in WHO clinical Stage IV ha the highest risk of mortality. The current stuy comes up with the same conclusion [4,17,19] except with stage II in t statistical significant from Stage IV. Results in [18] showe that HIV-TB co-infecte patients ha short survival time than negative in HIV-TB. The current stuy concurs with conclusions above. 5. Conclusions The stuy reveale that after initiation of the treatment, HIV-positive people ha estimate nearly 4 years meian survival time. From among the variables inclue in the stuy, four of them (Eucation level, marital status, stuy area an gener) i not have significant impact on survival. Resience, CD4 count, age, employment status, weight, WHO clinical stage, lymphocyte count an HIV- TB co-infection status ha significant impact on the survival experience of patients. Living in rural ha shortene the survival of patients. Patients, who were on job live longer compare to not working ue to illness patients. Higher boy weight i inicate longer survival, whereas low weight was associate with shorter survival. Oler age was associate with higher risk of eath. Higher CD4 cell counts were observe to have an association with better survival experience. Likewise higher lymphocyte counts were observe to have an association with better survival experience. HIV-TB co-infecte patients ha shorter survival time. An WHO clinical stage IV ha the highest risk of mortality. Mortality ue to AIDS was strongly associate with small amount or count of clinical characteristics, people living with HIV were often iagnose at a late stage an elaye initiation of treatment until they were severely ill or the clinical laboratory preictors reuce to small count. Promotion of testing services, which are wiely available in Ethiopia, is neee for earlier iagnosis an enrollment in treatment services an earlier initiation of ART with improve aherence counseling is in avance. Improvement of the clinical characteristics has strong association with longer survival of HIV patients, furthermore, further stuy require on the implication of ART service that brings improvement on clinical factors thorough the stuy perio. Acknowlegements We are grateful to the Wollo University for their collaboration an support the require fun in implementing HIV AIDS relate activities. We are particularly grateful to the Boru Mea, Dessie Referral Hospitals an Kombolcha Health Center aministrative an atabase teams an all the health workers, patient support groups an associations. References [1] Feeral HIV/AIDS Prevention an Control Office (HAPCO). Country Progress Report on HIV/AIDS Response, Feeral Democratic Republic of Ethiopia. 01. [] UNAIDS/WHO. AIDS epiemic upate UNAIDS 0 avenues Appia CH-111 Geneva 7 Switzerlan, 009. [3] USAID. HIV/AIDS Health Profile for Ethiopia. AA. Ethiopia Wikipeia, the free encyclopeia (008). List of countries by HIV/AIDS Ault prevalence rate. 010; Accesse on 30/05/10 [4] Oafe S., Ioko O., Baru T., Aiyenigba B., Suzuki C., Khamofu H., Onyekwena O., Okechukwu E., Torpey K. an Chabikuli1 O. N. Patients emographic an clinical characteristics an level of care associate with lost to follow-up an mortality in ault patients on first-line ART in Nigerian hospitals. 01; 15: [5] Office of AIDS Research Avisory Council (OARAC). Guielines for the use of antiretroviral agents in HIV-1 Infecte Aults an Aolescents [6] Unite Nations Programme on HIV/AIDS (UNAIDS). Report on the Global HIV/AIDS Epiemic, Geneva [7] Rogers, P., Sinka K., Molesworth A., Evans B., an Allarice G.. Survival after iagnosis of AIDS among aults resient in the Unite Kingom in the era of multiple therapies. Commun Dis Public Health, 001; 3: [8] Zahang C. HIV/AIDS relative survival analysis, M.Sc. thesis. Georgia state university, in the college of arts an sciences. 007: 79 p. [9] Tegiste Assefa an Eshetu Wencheko. Survival analysis of patients uner chronic HIV-care an antiretroviral treatment at Tikur Anbessa Specialize Hospital, Ais Ababa, Ethiopia. 01; 6 (1): -9. [10] Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. Journal of the American Statistical Association 1958; 53: [11] Cox D.R. Regression moels an life tables (with Discussion). Journal of the Royal Statistical Society 197; Series B, 34: [1] Hosmer DW, Lemeshow S, May S Applie Survival Analysis, n Eition. John Wiley & Sons, Inc., New York. [13] Aninet Worku Alemu an Miguel San Sebastián. Determinants of survival in ault HIV patients on antiretroviral therapy in Oromiyaa, Ethiopia. PMC, 010. [14] Johannessen A, Naman E, Ngowi BJ, Sanvik L, Matee MI, Aglen HE, et al. Preictors of mortality in HIV-infecte patients starting antiretroviral therapy in a rural hospital in Tanzania. BMC Infect Dis. 008; 8:5. [15] Sieleunou I, Souleymanou M, Schonenberger AM, Menten J, Boelaert M. Determinants of survival in AIDS patients on antiretroviral therapy in a rural centre in the Far-North Province, Cameroon. Trop Me Int Health. 009; 14: [16] Lee S. H., Kim K., Lee S. G., Cho H., Chen D. H., Chung J. S., Kwak I. S., an Cho G. J.. Causes of Death an Risk Factors for Mortality among HIV-Infecte Patients Receiving Antiretroviral Therapy in Korea. 013; 8 (7): [17] Gapayle A.K., Kumar N., Duggal A., Rewari B.B., Ravi V. Survival tren an prognostic outcome of AIDS patients accoring to age, sex, stages, an moe of transmission - A retrospective

7 7 International Journal of Data Envelopment Analysis an *Operations Research* stuy at ART centre of a tertiary care hospital. JIACM 01; 13 (4): [18] Alvarez-Uria G., Naik P. K., Pakam R. an Mie M. Factors associate with attrition, mortality, an loss to follow up after antiretroviral therapy initiation: ata from an HIV cohort stuy in Inia. Glob Health Action, 013; 6: 168. [19] Sieleunou I., Souleymanou M., Schonenberger A., Menten J. an Boelaert M. Determinants of survival in AIDS patients on antiretroviral therapy in a rural centre in the Far-North Province, Cameroon. Tropical Meicine an International Health, 009; 14: [0] Lutaloa T., Grayb R. H., Wawerb M., Sewankamboc N., Serwaa D., Laeyeneckere O., Kiwanukaa N., Nalugoaa F., Kigozia G., Nyanaboa A., Bwanikaa J. B., Reynolse S. J., Quinne T. an Openia P. Survival of HIV-infecte treatmentnaive iniviuals with ocumente ates of seroconversion in Rakai, Ugana. 007; 1 (6). [1] Rupali T. P., Mannam S., Bella A., John L., Rajkumar S., Clarence P., Pulimoo S. A., Samuel P., Karthik R., Abraham O. C., an Mathai D. Risk Factors for Mortality in a South Inian Population on Generic Antiretroviral. 01; 6.