Tuberculosis and human immunodeficiency virus co-infected patients’ mortality rate and its predictors in Dire Dawa, Eastern Ethiopia, 2018

INTRODUCTION

Tuberculosis and human immunodeficiency virus (TB/HIV) co-infection is a significant global public health problem. When people have HIV infection, and either latent or active TB disease considered as TB/HIV co-infected. Rapid growth of the human immunodeficiency virus (HIV) epidemic resulted in an equally dramatic rise in new tuberculosis (TB) cases. Which leads now a day’s co-infection of TB/HIV is common in many parts of the world, especially, in sub-Saharan Africa including Ethiopia.^1-3 The relation of the con-infection of TB and HIV is bidirectional. HIV infection is a leading risk factor for TB over-diagnosis of sputum smear-negative PTB and under-diagnosis of sputum smear-positive PTB. The HIV infection also threat for inadequate supervision of anti-TB chemotherapy, high morbidity and mortality during treatment, high default rates, high rates of TB recurrence and relapse, poor adherence due to pill burden.⁴ Similarly, TB increases viral load and replication. This fallouts in more rapid progression of HIV disease.⁴ The consequence of co-infection is multidirectional, started from the increasing of burden for the health workers to the loss of patients’ life.^5,6 Mortality rate of TB/HIV co-infected comes from many aspects or factors like weight, WHO staging, place of residence, CD4 count, age, types of and recurrence of TB, time of initiation of treatment.^7-10 Global efforts to combat TB/HIV co-infection have improved the mortality of patients after initiating TB and HIV treatment in different countries.^2,7,11-14 However in Ethiopia, few available evidence showed that insignificant decline of mortality rate and little predictors was known.^5,7,15 This study aimed to identify predictors and examine the mortality rate of the TB/HIV co-infected patients using retrospective cohort follow-up study.

Objective

The aim of this study was to determine TB/HIV co-infected patients’ mortality rate and its predictors in Dire Dawa, Eastern Ethiopia 2018.

METHODS

Study setting and period

Institution based retrospective cohort study design was used.

This study was conducted in the Dire Dawa town, Eastern of Ethiopia which is 515 km far from Addis Ababa. The patients were followed retrospectively from January 2012 to December 2016.

Setting: the study was implemented at both primary and secondary health institution; number of participating centres for follow-up were 471 individuals from those entered participants 244 were females and 302 participants were completed this study while 167 not completed (died, lost follow-up, or transfer out).

The baseline data were taken when the participants enter to the follow-up. The baseline data taken were sex, age, educational status, marital status, and residence, weight, WHO clinical stage, functional status, prophylaxis therapy status, other opportunistic infection status, hemoglobin count, CD4 count and the like variables were measured. While the secondary outcome were whether patient was died or survive in the follow-up period.

Populations

Source population

All TB/HIV co-infected patients treated in Dire Dawa public health institution from 1^st January 2012 to 31^st December 2016.

Study population

TB/HIV co-infected patients’ medical records in the period of 1^st January 2012 to 31^st December 2016 in the sampled institution.

Inclusion criteria

For this study All TB/HIV co-infected patients’ 1^st January 2012 to 31^st December 2016 were included. The inclusion criteria were being co-infected with TB/HIV and have started TB treatment.

Exclusion criteria

TB/HIV co-infected patients who have multidrug resistance. Patients’ records which has incomplete information and illegible handwriting, transfer in or truncation and interval censored patients were excluded from the study.

Sample size

Sample size for the survival status of TB/HIV co-infected patients followed to the anti TB treatment was used a two steps processes to determine the sample size the first one is determining the total number of expected event (death) and the second one is the total sample size required to get this number of events. Determined by the following consideration with a two-sided significance level of 5 percent and 80 percent power.

Were - M=is number of expected events (death)

- Z α⁄2=is the significance level two side (5% the value is 1.96)

- Z_β=80% power=0.842

-Ө is logarithm of the hazard ratio=9.1 (Abrha, et al).

- П is the fraction of subjects allocated to the first group (by general cohort study expected ratio of 1 to 3 exposed-non exposed ratio)

m=(1.96+0.842)/[Log (9.1)] ²*0.25×0.75

m=47

Where – n= is the number of subjects to be followed.

M= is the number of events (47)

P= s the over all probability of an event (death) at the end of the study= (20.2%).¹⁶

n=m/ P (events) = 47/0.202

n=280

The sample size for predictors of mortality rate of TB/HIV co-infected patients was calculated by Stata version 13 and one regression slope model by considering the following assumptions: confidence level 95%, power 80% and exposed to unexposed ratio of 1, standard deviation and correlation of covariates 0.5 and 0 respectively. From the calculated figures, taking the higher value that is 392 and add 20% of loss to follow up the final sample size will be 471 and taking it as finale sample size.

Data collection

Data were collected using a pretested check list which was adapted from different literatures. The check list has contained socio-demographic characteristics, TB related factors, HIV related clinical factors and drug related factors. Data were gathered from patients’ TB registers, ART registers, and laboratory requests, follow-up registration, pre-ART registries and clinical data were identified from medical file of patients for a five years follow-up.

Operational definition

Event: A TB/HIV patient who died from any cause after starting TB/HIV treatment.

Censored: patients who do not known a disease endpoint during their period of follow-up.

Survival status: time in days transpired from the date of initiation of TB treatment to death, lost follow up, transfer out or in the case of individuals who did not die (censored) the time in days transpired to complete follow up.

Patient and public involvement

In the process of this research informed consent were taken from all the heads of hospitals and health institutions. Patients’ directly involvement or direct contact with data measurer or data collector was not applicable since the data collection were from secondary document. However, the data collection process were realized with full consent of the institutions and confidentiality were fully appreciated. After the accomplished of the research the result were presented to the all institutions’ representatives and health berou officers. Moreover, it was presented to scholars of Haramaya University.

Data quality assurance

The collected data were checked, cleaned and compiled by the investigators as soon as data collectors reported on daily basis. If unfilled or missed data reported, the data collectors were asked to revisit and check the file the next day before proceeding to the next step. Finally, double data entry was done by two data clerks and consistency of the entered data was cross checked by comparing the two separately entered data.

Data processing and analysis

After the data were checked, entered into Epi-Data version 4.2.0.0.win.64, and exported to SPSS version 24 for analysis. The data were cleaned and edited before analysis. Data exploration was undertaken to see if there are odd codes or items that were not logical and then subsequent editing was made. The main end point or event in this study is death from TB/HIV co-infection. The response variable was mortality rate, defined as “time in days transpired from the date of initial TB treatment to death” or, in the case of individuals who were not die (censored), “the time in days transpired to complete study period, defaulter, lost follow-up, transfer out”. Variables were considered as a candidate if it shows marginal association (p<0.045) in the bivariate analysis. Multicollinearity was checked with standard error. Standard error (SE) less than 1.3 taken as non-collinearity and enter in to the multivariate analysis. Goodness of fit were checked by Schoenfeld residuals test and greater than 0.05 is taken as fitness of model or fulfilment of proportional hazard assumption. Time interaction were check by using cox time dependent covariance.

The patient cohort characteristics were described in terms of mean, median, standard deviations, and range values for continuous data; percentage, frequency tables and charts/graphs for categorical data.

For the comparison of time to recovery among the different groups of patients, Kaplan Meir curve were used. Log-rank test were used to compare survival curves of groups defined by categorical covariates. The Cox proportional hazard model were used to determine predictors of death after TB diagnosis. The crude and adjusted Hazard Ratio (HR) and its 95% confidence interval (CI) were estimated. Cox regression analysis were done with the identified candidate independent predictors.

RESULTS

Socio demographic characteristics of study participants

A total of 471 TB/HIV co-infected patients were followed for a median of 685 days from January 01/2012 to December 31/2016. The data were measured at baseline when patients enter to the study and followed till the end of study period for censored cases or till event happens. The mean age of the study subjects were 32.43 (SD±11.38) years with a majority of them lying between 15-45 years. Nearly one fourth, 106 (27%) had primary level of education. Almost two fifth study patients, 153 (40.5%), were self-employed (Table 1).

Clinical characteristics of the study subjects

Among study participants nearly three fourth, 340 (72.2%) were new TB cases. Four fifth of patients, 385 (81.74%), had weight between 30 to 60 kg. About 63.9% and 56.7% participants had >200 gm/mm³ CD4 counts and >10 gm/dl hemoglobin level respectively. Half, (50.5%) of study patients had PTB and from them 117 (24.8%) positive for AFB. More than 2/3^rd of participants, 327 (69.4%) initiated ART, out of them greater than half, 195 (59.5%), have initiated before 2 weeks of taking anti TB drug. Thirty-seven ART and 46 anti-TB patients were develop side effects. Approximately two fifth of case, 138 (37.5%) were had bedridden functional status. Concerning WHO clinical stages, 126 and 65 cases were in stage III and IV, respectively. Half of participants, 236 (50.5%) has started isoniazid and more than three forth, 360 (77.1%) started co-trimoxazole preventive therapy. Among all study subjects 93 (19.9%) developed other than TB opportunistic infection in study period (Table 2).

Mortality rate

Our hundred seventy-one patients followed for 344,972-person day observations (PDO) with median follow up of 685 days. A total of 79 (16.8%) TB/HIV co-infected patients were died during the study period with overall mortality rate 23 per 100,000 person-days of follow up. About half 40 (50.63%) of the deaths were occurred within the first 219 days of follow up and survival of 0.91 (95% CI: 0.88-0.93) after started anti-TB drug. Median survival time probability of patients were 0.82 (95% CI: 0.78-0.86). The overall survival rate of the patients were 83.2% (95% CI: 79.4-86.4) with 30 days minimum and 1822 days maximum survival rate of time length.

In this study participants who were infected by pulmonary TB 99 % more risk to death when compared to the extra pulmonary infected patients (log rank, x²=17.83, df (2), p<0.001) (Figure 1).

This study indicate that patients those not start CPT 2.45 times more likely risk to death as compared to those patients started CPT treatment (Log rank, x²= 20.25, df (1), p<0.0001) (Figure 2).

Based on this finding functional status was significantly related to the survival of patients (log rank, x2=22.53, df (2), p<0.001). Being bedridden six times, being ambulatory 4.2 times more risk to death to compared working patients at baseline respectively (Figure 3).

Predictors of mortality

Patients who were infected with pulmonary tuberculosis 99% more risk to death than extra pulmonary infected patients [AHR=1.99 (95% CI:1.16-3.41)]. The mortality risk among bedridden about six times [AHR=6.34 (95% CI: 2.43-16.59)] and ambulatory functional status four times [AHR=4.15 (95% CI: 1.57-10.98)] more likely to get death as compared to those working functional status. Regarding WHO clinical staging, patients who were in stage III and IV while starting anti TB treatment compared to stage I and II (combined), about three [AHR=2.88 (95% CI:1.56-5.30)] and four [AHR=4.20 (95% CI:2.21-8.01)] times more likely to have hazard of death respectively. In addition, patients who do not start CPT were more than two times more likely a risk to death as compared to those patients who already started CPT treatment [AHR=2.45(95%CI:1.54-3.91)] (Table 3).

*=(p<0.05), **=(p<0.001), CI=confidence interval, CPT=co-trimoxazole prophylaxis therapy CHR= crude hazard ratio, AHR=adjusted hazard ratios.

DISCUSSION

This study determined that the mortality rate and predictors of TB/HIV co-infected patients under anti-TB treatment. Near to 1 to 6 (16.8%) TB/HIV co-infected patients were died during the follow up of 344,972 PDO. Our finding was dissimilar with the studies conducted in Ethiopia, Ambo (15.77%) and Jimma (20.2%).^16-18

Malaysia 21.2%.¹³ However this finding was lower than studies conducted in Barcelona Spain (36.9 %), and Brazil (35%).^11,19 The difference could perhaps improvements in the health care systems and study time difference.

This study demonstrated that the PTB infection type significantly associated with the mortality of the TB/HIV co-infected patients and the finding come to an agreement with study conducted in the Ambo referral.¹⁷ This possibly will be related to meanwhile lung is vital organ delaying few days make patients susceptible to high hazard of death. The current study indicated that being ambulatory and bedridden functional status significantly affected the mortality of HIV/TB co-infected patients. This finding is consistent with other previous studies.^17,20 In this study, WHO clinical staging III and IV were 2.88 times and 4.2 times more likely at risk of dying compared to WHO clinical stages I and II. This finding is paralleled with the study conducted in Central and North West Ethiopia.^17,20 The possible explanation could be CD4 depletion, since patient’s body remains with poor defense and then opportunistic infections will get favorable condition to develop. Early initiation of cotrimoxazole prophylaxis has significant association with mortality of TB/HIV co-infected patients. This finding is in line with similar study conducted in North-West Ethiopia Bahir Dar town.²⁰

CONCLUSION

The findings of the current study illustrates that the mortality rate of TB/HIV co-infected patients were 23 per hundred thousand PDO. Therefore, special attention should be given for health education and early case detection. Supply of cotrimoxazole prophylaxis therapy is other alternative given attention. Interested researchers conduct prospectively and by increasing sample size to overcome the limitation of this study.

ACKNOWLEDGEMENTS

Authors would like to thanks Haramaya University, College of Health and Medical Science, for giving us this chance to did this study. Staffs of Haramaya university librarians and internet room coordinators for their cooperation. We also thanks dire dawa health birou, hospitals and health center workers.

Funding: No funding sources

Conflict of interest: None declared

Ethical approval: The study was approved by the Institutional Ethics Committee

REFERENCES

1. 1. WHO. Guide to Monitoring and Evaluation for Collaborative TB/HIV Activities--2015. 2015. Available at: https://www.who.int/publications/i/item/9789241508278. Accessed on 3 January 2023.
2. 2. WHO. WHO policy on collaborative TB/HIV activities: guidelines for national programmes and other stakeholders. 2012. Available at: https://www.who.int/publications/i/item/9789241503006. Accessed on 3 January 2023.
3. 3. Mekonnen D, Derbie A, and E. Desalegn, TB/HIV co-infections and associated factors among patients on directly observed treatment short course in Northeastern Ethiopia: a 4 years retrospective study. BMC research notes, 2015;8(1):666.
4. 4. Gebrezgabiher G. Treatment outcome of tuberculosis patients under directly observed treatment short course and factors affecting outcome in southern Ethiopia: a five-year retrospective study. PloS one. 2016;11(2):e0150560.
5. 5. WHO. Definitions and reporting framework for tuberculosis-2013 revision. 2013. Available at: https://apps.who.int/iris/handle/10665/79199. Accessed on 3 January 2023.
6. 6. Shaweno D, Worku A. Tuberculosis treatment survival of HIV positive TB patients on directly observed treatment short-course in Southern Ethiopia: a retrospective cohort study. BMC Res Notes. 2012;5(1):682.
7. 7. Federal H. AIDS prevention and control office. Country progress report on HIV/AIDS response. Addis Ababa: FHAPCO. 2014.
8. 8. Agbor AA. Factors associated with death during tuberculosis treatment of patients co-infected with HIV at the Yaoundé Central Hospital, Cameroon: an 8-year hospital-based retrospective cohort study (2006–2013). PloS one. 2014;9(12):e115211.
9. 9. Ferreira MD. Predictors of mortality among intensive care unit patients coinfected with tuberculosis and HIV. J Brasileiro de Pneumologia. 2018;44(2):118-24.
10. 10. Cain KP. Causes of death in HIV-infected persons who have tuberculosis, Thailand. Emerging infectious dis. 2009;15(2):258.
11. 11. Català L, Orcau A, De Olalla PG, Millet JP, Rodríguez-Mondragón A, Caylà JA et al. Survival of a large cohort of HIV-infected tuberculosis patients in the era of highly active antiretroviral treatment. Int J Tuberculosis Lung Dis. 2011;15(2):263-9.
12. 12. Vijay S, Kumar P, Chauhan LS, Rao SVL, Vaidyanathan P. Treatment outcome and mortality at one and half year follow-up of HIV infected TB patients under TB control programme in a district of South India. PLoS One. 2011;6(7):e21008.
13. 13. Ismail I, Bulgiba A. Predictors of death during tuberculosis treatment in TB/HIV co-infected patients in Malaysia. PLoS One. 2013;8(8):e73250.
14. 14. Teklu AM, Nega A, Mamuye AT, Sitotaw Y, Kassa D, Mesfin G et al. Factors associated with mortality of TB/HIV co-infected patients in Ethiopia. Ethiop J Heal Sci. 2017;27(1):29-38.
15. 15. Gesesew H. Predictors of mortality in a cohort of tuberculosis/HIV co-infected patients in Southwest Ethiopia. Infect Dis Poverty. 2016;5(1):109.
16. 16. Abrha H. Survival experience and its predictors among TB/HIV Co-infected patients in Southwest Ethiopia. Epidemiol (Sunnyvale). 2015;5:191.
17. 17. Refera H, Wencheko E. Survival of HIV-TB co-infected adult patients under ART in Ambo Referral Hospital, Ethiopia. Ethiop J Heal Develop. 2013;27(2):88-93.
18. 18. Abebe T, Angamo MT. Treatment outcomes and associated factors among tuberculosis patients in Southwest Ethiopia. Greatment Outcom. 2015;57(4):397-407.
19. 19. Dos Santos RP, Deutschendorf C, Scheid K, Goldani LZ. In-hospital mortality of disseminated tuberculosis in patients infected with the human immunodeficiency virus. Clin Develop Immunol. 2011;2011.
20. 20. Sileshi B. Predictors of mortality among TB-HIV Co-infected patients being treated for tuberculosis in Northwest Ethiopia: a retrospective cohort study. BMC infectious dis. 2013;13(1):297.