Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Evaluation of the TB-HIV Integration Strategy on Treatment Outcomes

477 views

Published on

Shared at a data dissemination and data use workshop on the results of the impact evaluation of the Strengthening Tuberculosis Control in Ukraine project. Access another presentation at https://www.slideshare.net/measureevaluation/evaluation-of-the-impact-of-a-social-support-strategy-on-treatment-outcomes/.

Published in: Health & Medicine
  • Login to see the comments

Evaluation of the TB-HIV Integration Strategy on Treatment Outcomes

  1. 1. Evaluation of the Impact Zulfiya Charyeva Nicole Judice MEASURE Evaluation, Palladium TB-HIV Integration Strategy on Treatment Outcomes
  2. 2. Strengthening TB Control in Ukraine Project (STbCU)  Goal – Reduce the burden of TB through specific quality assurance and system strengthening measures for routine TB services, MDR-TB, and HIV co-infection • Provide social support to promote patient adherence to TB treatment (social support study) • Improve access to and use of timely diagnostic and treatment for HIV co-infected patients to reduce mortality (TB-HIV integration study)
  3. 3. TB-HIV Integration Program Objectives  Identify gaps in TB-HIV co-infection services and build capacity to address them  Ensure HIV testing for TB patients and effective referral of those found to be HIV positive  Provide TB screening of HIV patients and referral to TB services for suspected TB cases
  4. 4. Activities Implemented by the STbCU Project  Work with the government to institutionalize best practices for TB-HIV management  Develop databases and protocols to support reporting and sharing of data across TB and HIV services  Provide numerous trainings to TB, HIV, and infectious disease (ID) specialists in caring for TB-HIV co-infected patients
  5. 5. Evaluation Design A mixed-methods approach with a quasi-experimental quantitative evaluation design complemented by qualitative descriptive work to inform the findings.
  6. 6. Impact Evaluation Questions: TB-HIV Integration Study  A. Completion of TB-HIV service cascade: What proportion of TB and HIV/AIDS patients complete each step in the cascade of services from screening to treatment per national protocol?  B. Factors affecting the use of TB-HIV services: What facilitates or impedes timely access to and use of testing and treatment for TB and HIV/AIDS patients?
  7. 7.  C. Impact of service integration on time to services: Do service integration, training and support between TB and HIV/AIDS services decrease the time lag between each step of service (screening, testing, and treatment) for TB and HIV/AIDS patients?  D. Impact of service integration on all-cause mortality: Do service integration, training and support between TB and HIV/AIDS services decrease all-cause mortality among the TB-HIV coinfected patients? Impact Evaluation Questions: TB-HIV Integration Study (2)
  8. 8. Summary of Methods, Table 1 Question Data collection Data sources Sample Sample size Analysis A, C, D Chart abstraction Patient medical records; electronic TB manager Systematic random sampling Baseline: 1,427 charts from facilities and 1,064 charts from AIDS centers. End line: 1,448 charts from TB facilities and 1,529 charts from AIDS centers. Survival analysis, proportional models with a difference-in- differences approach B In-depth interviews (IDIs) Patients, providers, STbCU staff Purposive Baseline: 18 IDIs with providers in six oblasts. End line: 30 IDIs with 17 IDIs and 6 focus group discussions with providers in 3 intervention oblasts, 6 IDIs with STbCU staff. Qualitative data analysis Context Facility survey Facility lead doctors and administrators All facilities in the regions Baseline: 18 TB and 9 HIV facilities. End line: 17 TB and 8 HIV facilities. Descriptive statistics
  9. 9. TB-HIV Integration Study Oblasts Intervention oblasts  Kharkiv, Odessa, and Zaporizhzhya  Selected based on TB and HIV case counts and co-infection rates Comparison oblasts  Kiev, Mykolaiv, and Zhytomyr  Loosely matched to intervention oblasts on TB and HIV disease rates, population density, and level of socio-economic development
  10. 10. Study Windows – Questions A, C, D  Baseline: January – December 2012  End line: April 2014 – June 2015
  11. 11. Sampling for Questions A, C, D TB facilities patient sampling:  First random sample (S1) of patients was selected without replacement from all new TB patients in the baseline/end line study window, proportionate to size of the oblast  A second sample (S2) was then selected from the remaining identified co-infected patients
  12. 12. AIDS centers patient sampling:  First random sample (S1) of patients was selected without replacement from the oblast AIDS centers registration journals in the baseline/end line study window, proportionate to size of the oblast  A second sample (S2) – the ID specialists in each oblast provided a list of all coinfected patients in the oblast • Systematic random sampling in Odessa • Use all remaining charts in other oblasts Sampling for Questions A, C, D
  13. 13. Difference-in-Differences Definition Source: Wikipedia, https://en.wikipedia.org/wiki/Difference_in_differences
  14. 14. Findings
  15. 15. RQA: Completion of TB-HIV Service Cascade – Findings from AIDS Centers
  16. 16. TB Screening and Testing Cascade for HIV Patients (Sample 1) – Figure 4.1
  17. 17. TB and HIV Treatment Cascade for HIV Patients (Co-Infected Patients) – Figure 4.2
  18. 18. RQA: Completion of TB-HIV Service Cascade – Findings from TB Facilities
  19. 19. RQA: Completion of TB-HIV Service Cascade – Findings from TB Facilities HIV testing:  In intervention oblasts, 91% of new TB patients with no prior HIV diagnosis received an HIV diagnostic test at baseline, compared with 99% at end line
  20. 20. 85% 82% 78% 7% 87% 86% 86% 14% 15% 5% 13% 5% 1% 93% 88% 88% 11% 92% 91% 91% 14% 7% 1% 8% 2% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% New TB patients HIV VCT HIV diagnostic test HIV case confirmed New TB patients HIV VCT HIV diagnostic test HIV case confirmed Baseline Endline Intervention No prior HIV Intervention Prior HIV Comparison No prior HIV Comparison Prior HIV HIV Testing Cascade for Newly Diagnosed TB Patients (Sample 1) – Figure 4.3
  21. 21. RQA: Completion of TB-HIV Service Cascade – Findings from TB Facilities ART initiation:  The percentage of HIV-positive TB patients with no prior HIV diagnosis that started ART increased over time in the intervention group from 21% at baseline to 51% at end line  ART initiation decreased from 48% to 47% in comparison oblasts
  22. 22. HIV Treatment Cascade for Co-Infected TB Patients – Figure 4.4 88% 35% 18% 18% 84% 56% 42% 42% 12% 3% 2% 2% 16% 8% 4% 4% 98% 78% 48% 48% 98% 71% 46% 46% 2% 1% 2% 1% 0% 20% 40% 60% 80% 100% HIV case confirmed HIV registration Started ART TB outcome recorded HIV case confirmed HIV registration Started ART TB outcome recorded Baseline Endline Intervention No prior HIV Intervention Prior HIV Comparison No prior HIV Comparison Prior HIV
  23. 23. RQB: Factors Affecting the Use of TB – HIV Services
  24. 24. Factors That Facilitate Access to and Use of Services  Improvements in timely TB diagnostic testing  Enhanced services for TB patients in AIDS centers  Tracking of TB clients who were successfully treated  Good communication between TB and ID specialists  Awareness of medical staff concerning HIV  Availability of free ART
  25. 25. Diagnostics became faster. New methods of sputum testing have appeared. Rapid tests for patients with co-infection. And Bactec and gin expert in case of TB … Informational support, laboratory diagnostics, methods of treatment – everything got systematized and improved. There has been integration of two services and by now we have pretty good services. [Focus group discussion participant] Factors That Facilitate Access to and Use of Services
  26. 26. … if the the patient has a fever, or if there are any other symptoms like cough, sweating, weight loss and etc., I immediately connect TB doctor. Thank God we have one in our facility. And, in general, it is very good, because, when there was no TB doctor, it was very difficult for us in this respect. And now, right here we can make a common decision whether to do a CT, or X-ray. [Provider] Factors That Facilitate Access to and Use of Services
  27. 27. Barriers to Timely Access to and Use of Services – Providers’ Perspectives  Clients’ inability to accept their HIV diagnosis and follow treatment instructions  Short-staffed facilities  Infrastructure issues
  28. 28. I came to work here in 2005 and the staff has not increased since that time, despite the fact that we have more and more patients. There should be 12 patients [per doctor], but in fact we have 36–40 [patients]. [Focus group discussion participant] Barriers to Timely Access to and Use of Services – Providers’ Perspectives
  29. 29.  Dealing with HIV-related stigma  Long lines at facilities  High out-of-pocket costs associated with travel, inpatient stay, laboratory work, and medications  Confusion about where to go to receive treatment  Confusion about medication regimens and their debilitating side effects Barriers to Timely Access to and Use of Services – Clients’ Perspectives
  30. 30. No, I don’t get the treatment by the place of my residence, but in the facility of XXX district. My treatment costs me a penny. I spend around 100 UAH only to get here and around three hours at my best, and I have to make as much as three transport changes. I have to travel to receive my treatment every day, which is very inconvenient. [Patient] Barriers to Timely Access to and Use of Services – Clients’ Perspectives
  31. 31. Barriers to Timely Access to and Use of Services Client databases are not consistently shared across all TB and HIV services  Makes coordination challenging  Further increases travel costs for patients, as they have to travel between TB and HIV clinics
  32. 32. RQB – Conclusion  The study suggests that while improvements in diagnostic testing and coordination across TB and HIV facilities is well underway, factors such as stigma, emotional burden, adequate education to deal with the side effects of the medication, and high patient out-of-pocket costs still need to be addressed.
  33. 33. RQC: Impact of Service Integration on Time to Services
  34. 34. RQC: Impact of Service Integration on Time to Services – Findings from AIDS Centers HIV Testing:  Patients in the intervention group were twice as likely at baseline (p<0.001) and 16% less likely at end line (p=0.115) to be tested for TB  Over the course of the TB-HIV integration program, TB testing improved significantly for both groups  In the intervention group relative to the comparison group, the net impact of the program on TB testing was negative (HR=0.40, p<0.001)
  35. 35. Figure 6.1: Time to TB Testing for Patients at the AIDS Centers (Sample 1) 0.000.250.500.751.00 0 200 400 600 800 1000 Time (days) Baseline 0.000.250.500.751.00 0 200 400 600 800 1000 Time (days) Endline TB testing among HIV patients Comparison Oblasts Intervention Oblasts
  36. 36. RQC: Impact of Service Integration on Time to Services – Findings from AIDS Centers ART initiation:  At baseline, patients in the intervention group were 37% less likely to begin ART compared to those in the comparison group (p<0.05) • This difference reduced at end line to 22%  The difference-in-differences model: the TB-HIV integration program resulted in a significantly positive impact on increase in ART testing in the intervention oblasts (HR=1.49, p<0.05)
  37. 37. Figure 6.2: Time to ART Initiation among Co-Infected Patients by Intervention Status Wald chi-square test: p= 0.292 0.000.250.500.751.00 0 200 400 600 800 1000 Time (days) Baseline 0.000.250.500.751.00 0 200 400 600 800 1000 Time (days) Endline ART initiation among co-infected patients by intervention status Comparison Oblasts Intervention Oblasts
  38. 38. RQC: Impact of Service Integration on Time to Services – Findings from TB Facilities HIV Testing:  TB patients in intervention oblasts were 42% less likely at baseline (p<0.001) and 23% less likely at end line (p<0.01) to be tested for HIV compared to TB patients in comparison oblasts  Difference-in-differences results model: a positive impact on the likelihood of receiving an HIV diagnostic test (HR=1.28, p<0.05)
  39. 39. Figure 6.3: Time to HIV Testing for Patients at TB Dispensaries (Sample 1)0.000.250.500.751.00 0 200 40 0 600 8 00 Time (days) Baseline 0.000.250.500.751.00 0 200 400 600 800 Time (days) Endline HIV testing among TB patients Comparison Oblasts Intervention Oblasts
  40. 40. RQC: Impact of Service Integration on Time to Services – Findings from TB Facilities ART initiation:  Patients in intervention oblasts were 53% less likely to initiate ART than patients in comparison oblasts at baseline (p<0.001), but were 35% more likely to initiate ART than the comparison group at end line (p<0.01)  Difference-in-differences model: a very strong and positive estimate of program impact on the likelihood of ART initiation (HR=2.91, p<0.001).
  41. 41. Figure 6.4: Time to ART Initiation among Co-Infected Patients at TB Dispensaries by Intervention Status0.000.250.500.751.00 0 200 400 600 800 1 000 Time (days) Baseline 0.000.250.500.751.00 0 200 400 60 0 800 1000 Time (days) Endline ART initiation among co-infected patients by intervention status Comparison Oblasts Intervention Oblasts
  42. 42. RQD: Impact of Service Integration on All-Cause Mortality
  43. 43. RQD: Impact of Service Integration on All-Cause Mortality – Findings from AIDS Centers  At baseline, there were no difference in survival between intervention and comparison groups  At end line, patients in the intervention group are about 14% less likely to die compared to the comparison group • The difference is not statistically significant  Difference-in-differences model: we do not detect a significant impact of the integration program on all-cause mortality
  44. 44. Figure 7.1. Time to death among coinfected patients at AIDS centers by intervention status
  45. 45. RQD: Impact of Service Integration on All-Cause Mortality – Findings from TB Facilities  No difference in the likelihood of all-cause mortality between intervention and comparison groups at baseline or at end line  Do not detect a significant program impact on the likelihood of death
  46. 46. Figure 7.2. Time to Death among Co-Infected Patients at TB Facilities by Intervention Status0.000.250.500.751.00 Proportionalive 0 100 200 300 400 500 600 Time (days) Baseline 0.000.250.500.751.00 0 100 200 300 400 500 600 Time (days) Endline Survival by intervention status Comparison Oblasts Intervention Oblasts
  47. 47. Conclusions
  48. 48. Conclusions Qualitative findings  The TB-HIV integration program affected several positive changes in the integration of services, especially around availability of diagnostic tests across facilities, and training of providers
  49. 49. Findings from HIV center records  The TB-HIV integration program is associated with a significant increase in timely initiation of ART Findings from TB facilities records  Significantly positive impact of the program on the likelihood of patients receiving a diagnostic HIV test and starting ARTs Conclusions (cont.)
  50. 50.  We do not detect an impact on survival based on data from either the TB or HIV facilities Conclusions (cont.)
  51. 51. Factors to Explain No Detection of Program Impact on Survival  At the time the patients entered AIDS centers, those in the intervention facilities might have been sicker  We were not able to account for disease severity variables such as CD4 cell count or TB disease stage in our impact models, due to the large amount of missing disease characteristic data at baseline, especially at AIDS centers
  52. 52. Questions and Discussion
  53. 53. MEASURE Evaluation is funded by the U.S. Agency for International Development (USAID) under terms of Cooperative Agreement AID-OAA-L-14-00004 and implemented by the Carolina Population Center, University of North Carolina at Chapel Hill in partnership with ICF International, John Snow, Inc., Management Sciences for Health, Palladium Group, and Tulane University. The views expressed in this presentation do not necessarily reflect the views of USAID or the United States government. www.measureevaluation.org

×