Novel Uses of Data in HIV/AIDS - an Informatics Perspective Rita Kukafka, DrPH, MA Assistant Professor of Public Health and Biomedical Informatics Columbia.

Novel Uses of Data in HIV/AIDS

- an Informatics Perspective

Rita Kukafka, DrPH, MAAssistant Professor of Public Health and Biomedical InformaticsColumbia University

May 8, 2003Washington, DC

An Informatics Perspective on Data

Informatics methods makes possible novel uses of data

Collect Once, Use Many Examples from HIV TIPS - data collected

once is used to:– Evaluate selected patient/provider

outcomes– Tailor patient education– Promote IT system use

What is HIV TIPS ?

Web-based clinical support system designed to improve antiretroviral treatment effectiveness

Combines informatics methods and behavioral science theory

Primary Outcome: Reduce suboptimal prescribing practices

The Problem: Information Overload for Providers

Two million facts needed for practice– 10,000 diseases and syndromes– 3,000 medications– 1,100 laboratory tests– 400,000 articles added to biomedical literature each

year– Responsible for approximately 1,000,000 facts – Medical literature doubling every 19 years– Doubles every 22 months for AIDS care

Information Overload for Patients

Google Search – HIV/AIDS – 1,660,000 hits– HIV/AIDS treatment – 619,000 hits– HIV/AIDS patient education – 134,000

People tend to believe what they read Exploitation of patients is common on

the Internet Patients trust their doctors

Solution

Apply technology to provideThe right informationIn the right formTo the right personIn the right placeAt the right time

Tailored

Information

Provider

Services

HIV TIPS (Functional Modules)

Web-based Interactive Guidelines– interactive guidelines made available over

the web, at the point of care Patient-Tailored Print

Communications (TPC)– a computer tailored patient report

designed to assist physicians in carrying out the counseling recommendations according to the guidelines

Functions of the Interactive Guidelines

1. Present medications from which to select medication plan

2. Review proposed medication plan3. Compare medication plan against guidelines4. Present message related to inadequate

therapy, drug interactions, and side effects of proposed medication plan

5. Create education plan for selected medications

6. Print education plan for selected medications

Functions of the Patient Tailored Adherence Education

1. Collect patient data related to adherence2. Process patient data related to

adherence to provide tailored education about improving adherence

3. Provide printed tailored patient education about improving adherence

Level of Assessment

Generic communication

Tailored communication

Interpersonalcommunication

Not assessmentbased

Based on Assessment ofindividuals

Con

ten

t of

Com

mu

nic

ati

on

gen

eri

cin

div

idu

ali

zed

Patient-Tailored Education

Rationale for Tailoring:

Unessential information is eliminated

Information is more personally relevant, people pay more attention

When information attended to and processed, it is more effective than nontailored information in helping the person enact desired behavioral changes

Tailored communication elicits..

• greater attention• greater comprehension• greater likelihood of discussion of the

content with others• greater change in cognitive-behavioral

mediating constructs addressed by the content

• greater likelihood of behavior change

HIV TIPS Patient Education Tailoring Process

Determinant specification

Tailoring questionnaire

Constructmessages

Data fileMessage

file4.Write algorithms/5.Automate tailoring

process

Tailored communications

Step 1

Step 2 Step 3

Steps 4,5

Tailoring Variables

Barriers: addresses important barriers that HIV-positive patients experience

Side Effects: side effects to HIV-Medication to provide pointers on how to deal with side effects that particular patient is reporting

Self-efficacy: six dimensionsManage mood, manage medication, manage symptoms, communicate with healthcare provider, get support / help, manage fatigue

Medication review: provides adult dosing, possible side effect, notes and medicine image to the selected drug

System Architecture

HIV/AIDS Health care provider’s web browser

Web SeverCGIDB Driver

Internet

Service ProviderWeb browser

KnowledgeDatabase

nyc2000

******

HIV TIPS Innovative Approach

Designed for community providers without sophisticated technology

Maximize flexibility – 42 agencies with no uniformity of practice patterns

Use case scenarios used design navigation

Individual and agency level needs assessment to promote system use

30 – 70% IT projects fail – not adopted by end user

Agency Level Needs Assessment

Collect information on the technology resources, staff readiness, task flow, and barriers in IT use in the setting

Module developed to instruct AETCs on how to perform the assessment

A participatory design process to plan how the HIV Tips can be best used in the setting

If necessary, the clinic will be linked with appropriate resources to increase capacity

Provider Level Assessment

Collect provider-level data relevant for staging readiness to use the HIV TIPS software

Design provider-level tailored information and education material to motivate use

Use of push technology (email) to deliver provider tailored materials

Conceptual Diagram ofEnhanced Intervention

Usage Models(scenarios specifying roles for different users in a clinic)

Theory of Participatory Design

General Characteristics of Clinic (Capacity Survey, Observation Log, Attitude Survey)

Technology Inventory

Provider Attitudes (Staging Questionnaire)

Task Flow Assessment

Barriers Assessment

-Needs assessment data summarized in terms of barriers

Participatory Design Meeting (HIV TIPS Implementation Map)

Evaluation of Participatory Design Process:(Participant Survey, Design Process Description Tool)

Ongoing Support to Clinic

Computerized Matching of Staging with Tailored Messages

Tailored E-mail Messages to Providers

Ongoing Support to Users

Transtheoretical Model (Stages of Readiness)

GUIDING CONCEPTS

GUIDING CONCEPTS

ORGANIZATION LEVEL

INDIVIDUAL PROVIDER LEVEL

Evaluation of the HIV TIPS Intervention

Group Randomized Assignment of Clinics into 3 Treatment Conditions

• Enhanced IT Intervention• IT Intervention• No Intervention (control group)

Intervention Activities,for each Treatment Condition

Treatment Conditions →

Activities ↓

EnhancedIT

InterventionIT

Intervention

(control group)

No Intervention

IT Needs Assessment

access to HIV TIPS motivational and informational

Tailored Communications

Agency Recruitment

Agencies to be Recruited from large group participating in AETC longitudinal training initiatives.

12 Agencies from the Midwest AETC

30 Agencies from NY/NJ AETC

Study Selection Criteria

for Agencies/Clinics:– 25+ HIV-positive patients– Internet Access

for Patients:– Adult patients (non-pediatric,

not pregnant)– 2+ visits over past 12 months

(1 per 6-month period)– CD4 Counts < 500 or on

Medications

Provider-Level Variables of Interest

Exposure to Intervention:– Frequency with which HIV TIPS is accessed– Frequency that different components are used– Provider satisfaction with HIV TIPS

Patient-Level Variables of Interest

Descriptive/Demographic Information

– gender– age– race/ethnicity– housing status– insurance coverage– substance abuse/mental health

Current Health Status– CD4-cell count– HIV viral load– current opportunistic infections– current adverse drug reactions

Current Anti-Retroviral Regimen

– drugs prescribed– Dosage– dosing schedule– any change in medications

Medication History– previous anti-retrovirals taken– previous adverse drug reactions– viral resistance testing, if available– other medications that have known

HIV/AIDS-drug interactions

Sources of Evaluation Data

Treatment Conditions →

Data Sources ↓

EnhancedIT

Intervention

IT Interventi

on

No Intervent

ionclinic & training information from

AETC Information Forms

frequency and extent of IT use

from Website Usage Data

satisfaction with IT gauged in

Provider Surveys

Changes in provider readiness (barriers to

system use)

Changes in patient variables associated with

adherence

client demographics, health status, & medications from

Chart Review

Stored data- used to tailor messages Stored data – used to tailor messagesextended design

Summary

Examples from HIV TIPS – collect once, use many – tailoring patient education– tailoring provider to increase system usage– evaluation

Informatics methods (e.g., data storage, retrieval, sharing) makes optimal the use of biomedical information, data, and knowledge

As illustrated by HIV TIPS can be used for problem solving and decision making in addition to evaluation