People with HIV providing high quality treatment support in community

A study conducted in rural Kenya demonstrates that task-shifting HIV treatment programmes from clinic staff to community health workers, specifically those living with HIV, is both feasible and acceptable, giving support for further exploration of this model and providing alternatives to the physician-centred approach.

The two-year long study, published in the September edition of the Journal of the International AIDS Society, was conducted by American and Kenyan investigators at a United States Agency for International Development-Academic Model Providing Access to Healthcare (USAID-AMPATH) Partnership clinic in central Kenya.

Task-shifting has been advocated by the World Health Organization and other health experts in the context of a clinical health care worker shortage in resource-limited settings.

The study evaluates a community care coordinator (CCC) model, in which HIV-positive care workers provide support for their clinically stable HIV-infected peers with the assistance of personal digital assistants (PDAs). Results show clinic visits being reduced by 50% among the intervention group and accepted by clinic staff, patients, and CCCs themselves. CCCs were also able to identify patients' psychosocial problems and monitor treatment adherence.

The USAID-AMPATH Partnership clinic Mosoriot Rural Health Centre, located in the rural Kosirai Division of Kenya, was the site of the study. Eight of the 24 sub-locations of the Kosirai Division were randomly assigned to the intervention, with the rest acting as controls.

Nine HIV-positive patients from the Mosoriot clinic were selected to be CCCs, underwent didactic and practical training, and were paid an outreach worker's salary, lower than that of clinical staff. Eight entered the field, with one acting as an alternate. All eight remained in the field for the duration, managing between eight and 20 patients each.

In the CCC model patients were seen every three months at the clinic rather than monthly as occurs in the Mosoriot standard of care model.

During the two-month interim, patients would be visited monthly by their CCC, who measured their vital signs and reviewed symptoms as guided by a preprogrammed PDA: CCCs would ask patients a series of questions regarding their health, such as the existence of a cough or vomiting, and record answers in the PDA. If a patient displayed a symptom or collection of symptoms, the device prompted further questions, such as their length and severity, and triggered an alert if patients' answers fell outside of pre-established parameters, giving detailed instructions to CCCs for further action. CCCs also dispensed patients' monthly supply of ART and opportunist infection medication.

CCCs underwent continuous evaluation and assessment throughout the study - more rigorously during the mentoring period - and were given superior, satisfactory, or unsatisfactory evaluation scores by clinical officers.

At the end of the first year, 133 formal evaluations had been completed on the eight active CCCs (16 to 17 evaluations per CCC). CCCs consistently received superior summary scores: 89% of all summary scores were superior and the remaining satisfactory. 88 evaluations (11 per CCC) were undertaken in Year 2, during which 94% of summary scores were superior. Clinic staff also monitored CCC performance by comparing data collected at home visits, including that recorded in PDAs, with data collected at the clinic.

Investigators used evaluations and transcripts of monthly CCC meetings to assess the model, highlight and address problems, and improve the programme throughout the study. Investigators also regularly met with CCCs and clinical officers to discuss barriers and enhance CCC performance.

At the trial's end, of those not lost to follow-up, 64% (56 of 87) of patients in the intervention arm and 52% (58 of 103) in the control arm were willing to continue (p = 0.26). Additionally, CCCs were contacted by non-enrolled patients who wanted to receive such care. As such, investigators claim that patients accepted the programme.

In addition to reducing patient visits, the CCC model resulted in greater identification of and support in resolving psychosocial issues, such as alcohol abuse and food insecurity, than in the standard of care model. CCCs themselves also felt that they were better able to monitor treatment adherence than clinic staff as pills could be less easily hidden during home visits. Finally, CCCs acted as a link between the AMPATH pharmacy, outreach and clinical teams, and patients.

The author's warn that such a programme is not cost free. Training, PDAs, and salaries must be accounted for, but may be partially offset by the decrease in clinic visits.

Additionally, HIV disclosure remains an issue for patients, and therefore similar programmes should aim to avoid the "AIDS label."

It took longer than anticipated for CCCs to adapt to new technologies such as PDAs, and therefore more time should be given to training with such instruments; and mechanisms must be put in place to facilitate CCC referral to the clinic and clinic referral of follow-up of particular issues to the CCCs.

Issues such as stigma and client and staff expectations were highlighted and addressed throughout the study's period. The authors also stress the importance of finding committed individuals to act as CCCs.