Study Results
CASIC Builder™ evolved from our study done
under NIH project SBIR #2R44AI44558-02A1, our phase II SBIR study.
This study was performed in conjunction with the Veteran's
Administration Query-HIV
program. Results were presented at the
American Conference for the Treatment of HIV in Atlanta, GA, Sept.
15-18, 2005 by Allen Gifford, MD, Boston University School of Public Health,
Department of Health Policy & Management.
The conference poster
describes background and objects, study methods, study results. This information is summarized below.
Phase II study results summary
- HIV/Adhere audio-CAPI is accepted, feasible and clinically useful.
- Patients who are making significant medication errors and who are non-adherent
to ARV's can be identified.
- Nonadherence was more often missed by providers in older, and more
educated nonadherent patients.
- Tailored adherence counseling messages and patient reasons for missing
doses may help providers communicate with patients about ARVs.
- Depression and substance abuse are often detected using HIV/Adhere screening.
In our initial phase I study, we developed a pilot version of HIV/Adhere.
The study researched the ability the software's ability to detect patient
medication misunderstandings and nonadherence. One hundred eleven patients and
eleven providers participated in the study. The results from this initial effort were published here:
Bangsberg, DR, Bronstone, A, Hofmann, R. A
computer-based assessment detects regimen misunderstandings and nonadherence
for patients on HIV antiretroviral therapy. AIDS Care. 2002;14:
3-15.
Phase I study results summary
- Over half of the patients who completed the assessment made at least one error in describing their ARV medication regimen, with 14% failing to name at least one of their medications and 14% identifying a medication not currently prescribed.
- 30% of patients reported taking less than 90% of their ARV medications over a 3-day period. This proportion of nonadherent patients is comparable to the proportions identified in recent studies of HIV-seropositive individuals.
- Comparison of provider estimates and patient reports of adherence revealed that providers tend to overestimate their patients' adherence, correctly identifying only 24% of nonadherent patients.
- Patients' reports of adherence on HIV/Adhere were significantly associated with viral load counts, supporting the validity of self-reported adherence.
- The vast majority of patients found HIV/Adhere easy to complete, and reported that it accurately reflected they way they take their medicines.
- Seventy-five percent of patients felt that it was important for their provider to see their assessment results, and most providers felt that the Patient Adherence Reports were an extremely valuable aide in identifying patients' misunderstanding of their regimen and nonadherence.
- The act of completing the assessment in itself was a potentially powerful intervention, as 65% of patients indicated that it made them think more about how they take their medicines.
Participating providers were interviewed after receiving the Patient Adherence Reports for their patients who participated in the study. Below, we present direct quotations from providers exemplifying the four major themes which emerged from the interviews:
- Providers were often surprised by the information revealed in the Patient Adherence Reports
- Providers believed that patients are more likely to disclose nonadherence to the computer than to the provider
- Providers do not have enough time to conduct thorough adherence assessments in their practices, and
- Providers believe that incorporating computerized patient medication assessments into their practice would help them be more efficient and effective.
Providers were surprised by information revealed in the Patient Adherence Reports.
"Now, here's another good thing. He says here…he's taking Sustiva, which
I had him on. And he also said he's taking Nevirapine, which I didn't
have him on….Now, I don't understand that. Obviously, that brings up a
problem….Probably he was on one and it was discontinued and he never stopped
[taking] it."
"This actually helps…I can't figure out why his liver is shot. Now I
say, 'Ah, ha! He's taking six pills instead of two pills of the drug that's
worst on the liver.'….So this is a good example of where this actually
helps."
"Look at this…he's missed all of his medications. He's on a drug holiday
that I didn't know about."
Providers believed that patients are more likely to disclose
nonadherence to the computer.
"Why are people more honest with the computer than they are with the
doctor? If I ask him if he misses any doses, he says, 'Oh, no.' Then I
look at this [the report], and he's only taking [his medication] 1/3 of
the time…he's supposed to be on Norvir and he's missed every single dose…yet,
you ask him and [he says] everything's fine."
"People can tell anything to the computer without feeling any judgement
or attitude back. I can imagine people feeling comfortable to 'fess up'
[to the computer] where they wouldn't necessarily to their clinician who
has expectations of them…the interaction is with the machine and I think
that could make people more open.
Providers lack the time to conduct thorough medication assessments.
"I don't sit down and ask them about every single drug in this detail.
Basically, I say, 'Are you taking your medications?' 'Have you missed
any doses?' But I don't go and ask them, 'how many Sustiva have you missed?
How many Nevirapine have you missed?' So this helps."
" I usually ask what they are taking and most of the time how many times
a day. But I don't always ask everybody what medications are you taking,
how many times a day, how many pills."
Providers believe that computerized patient assessments can save
them time and provide valuable information.
"I would put this [adherence report] in the chart and give a copy to
the patient so they would have a reference. I think that would probably
cut down on some of our phone calls."
"When I change a cocktail in somebody, I typically will write the prescription
and then write all this stuff [dosing instructions] and it takes 8 minutes
just to do that. This could be very helpful to develop a program and push
a button. That is how I like things."
"It seems the more patients can do on their own and then have communicated
to us, the better."
Please visit our publications
page for references to additonal studies that have used HIV/Adhere.