Factors Associated with Loss to Follow-up and Mortality in Public - - PowerPoint PPT Presentation

Factors Associated with Loss to Follow-up and Mortality in Public Sector MDR-TB Treatment in Western Maharashtra, India from 2015-2016

Matthew Murrill,1 Sudhakar W. More,2 Sanjiv W. Kamble,3 Vaishali Kulkarni,4 David Dowdy,5 Amita Gupta1

1Johns Hopkins School of Medicine, Baltimore, MD, USA 2State TB Training and Demonstration Centre, Pune, India 3State TB Office, Maharashtra, Pune, India 4DR TB Centre, Aundh Chest Hospital, Pune, India 5Johns Hopkins Bloomberg School of Public Health, Epidemiology, Baltimore, MD, USA

[X] I have the following conflict of interest(s) to report: [X] Receipt of grants/research supports

• NIH Medical Scientist Training Program
• UJMT Fogarty Global Health Fellowship

[ ] Receipt of honoraria or consultation fees [ ] Participation in a company sponsored speaker’s bureau [ ] Stock shareholder [ ] Spouse/partner [ ] Other

Conflict of interest disclosure

Study Setting: Western Maharashtra

Methodology

• Limitations of prior work:

(1) specific risk factors absent, (2) composite outcomes, (3) competing risks

• Study design: retrospective cohort analysis
• Study population: individuals starting MDR TB treatment at Aundh Chest

Hospital from 5 districts of Maharashtra between Jan 2015 to Dec 2016

• Outcomes: loss to follow-up (≥2 months), death (from any cause)
• Risk factors: clinical, demographic, health service, social
• Analysis: survival analysis (subdistribution and Cox proportional hazards)

Methodology: Data Sources

• Treatment cards (DOTS provider)
• Treatment cards (TU)
• Pre-treatment evaluation: medical records
• Treatment cards (DR TB Centre)
• Treatment register (DR TB Centre)
• Treatment cards (DTO)
• Treatment register (DTO)
• Adjudication with DOTS-Plus Supervisors

Peripheral Health Institution (PHI) Tuberculosis Unit (TU) District TB Office (DTO) DR-TB Centre

Factors Associated with Loss to Follow-up and Mortality (n = 684)

Conclusions

• Primary findings:
• Treatment outcomes similar to

national averages

• Loss to follow-up: alcohol use,

prior private treatment

• Mortality: undernutrition, anemia
• Implications:
• Highlights potential of leveraging

existing public sector data

• Need to integrate alcohol

treatment interventions into care

• Strengths:
• Leveraged multiple programmatic data sources
• Competing risks survival analysis
• Challenges:
• Data sources:
• Data standardization and integration
• Missing data
• Outcome status:
• Loss to follow-up tracing efforts not done
• Transfer out, switch to XDR TB treatment

were censored

• Public sector treatment data only

Acknowledgements

• Revised National TB Control Programme, Maharashtra
• Dr. Sanjiv W Kamble (State TB Office)
• Dr. Sudhakar More (STDC, Pune)
• Ms. Vaishali Kulkarni (Aundh Chest Hospital)
• City TB Officers and DOTS-Plus Supervisors
• Johns Hopkins University SOM and SPH
• Dr. Amita Gupta
• Dr. David Dowdy
• Dr. Vidya Mave, Dr. Nishi Suryavanshi
• Dr. Nicky Mehtani
• Ms. Pallavi Kadam, Mr. Mahesh Madane
• Funding and support: NIH Medical Scientist Training Program (JHU), UJMT Fogarty Program