Telemedicine was introduced in the U.S. in the late 1950s using the limited technology of the time, and despite enormous technological advances that have made remote real-time audiovisual communication, examination, and robotically controlled surgical procedures possible, the practical application of telemedicine remains limited. Provider acceptance remains a significant factor in the limited use of telemedicine, which is a particular concern for HIV-treating physicians since a satisfactory/trusting patient/provider relationship has been identified as one of the most significant factors in predicting adherence to antiretroviral therapy.⁽¹⁾ In prisons in particular, where there are so many other potential barriers to effective HIV care, introducing a new and untried technology may seem particularly imprudent to providers.

Since nonadherence is the primary reason for antiretroviral failure and the development of viral resistance (and cross-resistance)⁽¹⁾ to the growing, but still limited number of antiretroviral agents, it is imperative that the effectiveness of new treatment modalities be questioned. Unfortunately, clinically significant medical outcome studies on the effectiveness of telemedicine in delivering HIV care have not been reported. Additional concerns include the rapid obsolescence of telemedicine equipment, and some providers have had difficulty furnishing complete paper record copies to telemedicine receiving sites. However, implementation of electronic health records is on the rise and should circumvent the latter issue. Another significant factor may be the substantial cost of T-1 lines, even if in fractional use. Recent dissemination of complementary, lower cost technologies may decrease telemedicine costs. Some published reports have demonstrated that telemedicine has reduced correctional health care costs, which is attributed to a reduction in security and other travel-associated costs incurred when incarcerated patients are transported from remote sites to specialists.⁽²⁾

In late 1997, TDCJ Correctional Managed Care developed and implemented a telemedicine program aimed at providing specialty care from a site on the University of Texas Medical Branch at Galveston (UTMB) campus to patients incarcerated in over 90 prison units in eastern Texas. For HIV care, as well as other specialty care, this was a major shift in health care delivery since HIV positive offenders had always been transported to Hospital Galveston (HG) for HIV care and participation in clinical trials at UTMB.

Despite the fact that a busy eight hour/week telemedicine clinic with an average census of 35 to 40 patients per session has been established, an on-site HG HIV Clinic has been maintained to deliver health care to patients for whom telemedicine has been deemed inadequate. It is the clinical judgment of the telemedicine provider to determine whether telemedicine is inadequate. There are no pre-established criteria. The usual reasons for requesting an on-site visit rather than a telemedicine follow up are that the provider does not feel that (s)he is "connecting" with the patient, there is a physical exam finding that cannot be evaluated using telemedicine, or the patient's presentation is too complicated and needs extra attention in an on-site visit. Failure for a patient to achieve a nondetectable virus load is not a reason for an on-site visit as long as the provider feels that (s)he is adequately communicating with the patient and the patient does not have a physical exam finding that requires a hands-on exam. Additionally, UTMB's AIDS Care and Clinical Research Program maintains a clinical trials program for HIV-positive prisoners, and those who wish to be screened and followed on clinical trials must travel to Galveston for initial screening and study-related follow up. The introduction of telemedicine provided an opportunity to compare telemedicine consultation with on-site (HG) consultation in the HIV care program, and to determine whether HIV care (both telemedicine and HG on-site care) was comparable to local community standards. This study was carried out by conducting a chart review.

As part of the assessment, we asked the question, is there a difference in the proportion of patients with a nondetectable virus load seen in the community clinic staffed by the same providers who staff the TDCJ HIV clinics. A random sample of charts of patients seen between September 2000 and March 2001 was selected from the Community Clinic (unincarcerated patients), TDCJ Telemedicine Clinic, TDCJ HG Clinic (TDCJ patients for whom telemedicine has been deemed inadequate), and TDCJ Research Clinic (HG clinic for incarcerated patients participating in clinical trials) and were reviewed for pertinent clinical information including the proportion of patients receiving antiretroviral treatment and most recent viral load determination. Since incarcerated patients may attend any of the three TDCJ clinics depending on clinical circumstances or clinical trial participation, charts were included for review only if the last three visits were conducted in the same clinic, or if there were only two documented visits, both must have occurred in the same clinic (see Table 1).

Most patients were seen multiple times during the six-month period covering this chart review. Therefore, the approximate number of patients seen in each clinic can be determined by dividing the number of appointments by three. The number of charts reviewed represents a 15 to 20% sample of the total number of patients seen in each clinic.

A one way analysis of variation (ANOVA) and binomial confidence limits analyses were performed in order to analyze the viral load response data. The results showed that a significant difference cannot be detected between the viral load responses in Group 1 (Community), Group 2 (Telemedicine), Group 3 (Hospital Galveston), and Group 4 (HG Research). Although there was a trend towards a significant difference in the HG Research Group, this reached borderline significance in the binomial confidence limits analysis only. These results are limited.

This chart review was not a scientifically designed study; it was done as part of an internal audit program to determine how to improve HIV care in the TDCJ. Patients were not matched for disease stage, sex, race, antiretroviral history, or other factors, which might influence viral load response, but the samples were randomly selected.

The next step is to conduct an appropriately powered, prospectively-designed trial that considers variables that determine the overall effectiveness of HIV health care delivery.

Yet, some preliminary conclusions can be drawn. First, this information agrees with other reports^{(3, 4)} indicating that while approximately 80% of clinical trial participants (HG research) have a reduction in viral load to below the levels of detection, the responses in clinical practice are more variable. This finding is expected since relatively well patients who are less likely to have drug-resistant viruses and who have been selected for their ability to adhere to medical treatment should have the best response to investigational antiretroviral regimens. Second, the viral load responses of patients seen by ACCRP HIV Specialists are similar whether the patients are seen by telemedicine or in traditional face-to-face encounters. These preliminary results are very encouraging. We interpret the results to reflect our experience that telemedicine can result in the type of patient/provider relationship that translates into adherence and reduction in viral load to nondetectable levels in a select group of patients.

References

1. HIV Treatment Guidelines available at http://www.hivatis.org.

2. Paar, David, Telemedicine in Practice: Texas Department of Criminal Justice, HEPP News, May 2000.

3. S. Guerin, L. Meyer, I. Thodorou, F. Boufassa, M. Magierowska, C. Goujard, C. Rouzioux, P. Debre, J.F. Dlefraissy, and SEROCO/HEMOCO Study Group. Deletion and Response to HAART in HIV-1-Infected Patients. 7th Conference on Retroviruses and Opportunistic Infections. Abstract 452, 2000.

4. B.Y. Nguyen, D.W. Haas, C. Ramirez-Ronda, M.A. Thompson, J. Gallant, J. Currier, D. Paar, C. White, A. Collier, D. Mehrotra, M. Chung, C. Harvery, J. Chodakewitz, Thirty Two Week Follow-up of Indinavir Sulfate (IDV) Administered Q8 Hours (H) Versus Q12 H in Combination With Zidovudine (ZDV) and Lamivudine (3TC): The 5th Conference on Retroviruses and Opportunistic Infections. Abstract 374, 1998.

Back to the HEPP News June/July 2001 contents page.