Reactive Adherence Counseling (Haiti)

I-Tech, Haiti (April 2021)

Countries: Haiti

Intervention Description

Introduces predictive analytics to improve HIV clinical management. It pairs improvements to a routine data system with theory-based behavior change interventions for patients and health care workers.

Presenter: Nancy Puttkammer, I-TECH

DUC Meeting Recording

Intervention Details

Intervention Attributes	Text
	· Appointment Date (history) · Visit Date · Proportion of days covered based on dispensing (changed with COVID) · Sex · Age · Marital status · Time from HIV diagnosis to enrollment in HIV care and treatment · Time from HIV diagnosis to ART initiation · Baseline CD4 count · Body mass index · World Health Organization (WHO) stage of HIV disease progression ART regime.
	Puttkammer N, Simoni JM, Sandifer T, et al. An EMR-Based Alert with Brief Provider-Led ART Adherence Counseling: Promising Results of the InfoPlus Adherence Pilot Study Among Haitian Adults with HIV Initiating ART. AIDS Behav 2020. PMID: 32715409
	Can work in EHR Interest in exchange of risk information with community health worker tool/app
	Risk level using data elements at left
	Need to implement within OpenMRS-based EMR Prediction model needs refreshing as conditions change (multi-month dispensing)
	Most helpful when using EHR at the point-of-care Will need to be re-implemented with OpenMRS rollout Threshold of data quality needed (being investigated) Prediction models are being revisited using machine learning (super learner models) with additional data elements.
	Complexity of implementing a clinical decision support prediction model needs to be weighed against the benefit of added sensitivity and specificity gained (is it worth it) Need governance (what sensitivity and specificity of prediction is adequate; whether it’s suitable to implement in all facilities using EMR or only sites with sufficient data quality) Need actionable clinical and psychosocial support for those determined to be at high risk

Intervention Attributes

Text

· Appointment Date (history)

· Visit Date

· Proportion of days covered based on dispensing (changed with COVID)

· Sex

· Age

· Marital status

· Time from HIV diagnosis to enrollment in HIV care and treatment

· Time from HIV diagnosis to ART initiation

· Baseline CD4 count

· Body mass index

· World Health Organization (WHO) stage of HIV disease progression ART regime.

Puttkammer N, Simoni JM, Sandifer T, et al. An EMR-Based Alert with Brief Provider-Led ART Adherence Counseling: Promising Results of the InfoPlus Adherence Pilot Study Among Haitian Adults with HIV Initiating ART. AIDS Behav 2020. PMID: 32715409

Can work in EHR

Interest in exchange of risk information with community health worker tool/app

Risk level using data elements at left

Need to implement within OpenMRS-based EMR

Prediction model needs refreshing as conditions change (multi-month dispensing)

Most helpful when using EHR at the point-of-care

Will need to be re-implemented with OpenMRS rollout

Threshold of data quality needed (being investigated)

Prediction models are being revisited using machine learning (super learner models) with additional data elements.

Complexity of implementing a clinical decision support prediction model needs to be weighed against the benefit of added sensitivity and specificity gained (is it worth it)

Need governance (what sensitivity and specificity of prediction is adequate; whether it’s suitable to implement in all facilities using EMR or only sites with sufficient data quality)

Need actionable clinical and psychosocial support for those determined to be at high risk

PreviousReactive Adherence Counseling Interventions NextAdherence Dashboard (Kenya)

Last updated 1 year ago

Intervention Attributes	Text
Data Elements	· Appointment Date (history) · Visit Date · Proportion of days covered based on dispensing (changed with COVID) · Sex · Age · Marital status · Time from HIV diagnosis to enrollment in HIV care and treatment · Time from HIV diagnosis to ART initiation · Baseline CD4 count · Body mass index · World Health Organization (WHO) stage of HIV disease progression ART regime.
Evidence	Puttkammer N, Simoni JM, Sandifer T, et al. An EMR-Based Alert with Brief Provider-Led ART Adherence Counseling: Promising Results of the InfoPlus Adherence Pilot Study Among Haitian Adults with HIV Initiating ART. AIDS Behav 2020. PMID: 32715409
Technology Requirements / Interoperability	Can work in EHR Interest in exchange of risk information with community health worker tool/app
Calculations / Algorithms	Risk level using data elements at left
Factors to Scale	Need to implement within OpenMRS-based EMR Prediction model needs refreshing as conditions change (multi-month dispensing)
Implementation Considerations	Most helpful when using EHR at the point-of-care Will need to be re-implemented with OpenMRS rollout Threshold of data quality needed (being investigated) Prediction models are being revisited using machine learning (super learner models) with additional data elements.
Governance Considerations	Complexity of implementing a clinical decision support prediction model needs to be weighed against the benefit of added sensitivity and specificity gained (is it worth it) Need governance (what sensitivity and specificity of prediction is adequate; whether it’s suitable to implement in all facilities using EMR or only sites with sufficient data quality) Need actionable clinical and psychosocial support for those determined to be at high risk