An analysis of decision-making by non-specialist physicians, clinical officers and nurses in Uganda has found that there is a high level of agreement in decisions on starting antiretroviral therapy.
The study, in the journal Human Resources for Health, contributes to the evidence base in support of task-shifting, and may ease concerns about a decrease in quality compared to current standards of care. It lends support to arguments for increased investment in training of nurses and clinical officers for the delivery of antiretroviral therapy in rural and semi-rural settings.
Rapid scale up of antiretroviral therapy has brought to light weaknesses in the health systems of developing countries. WHO estimates that over four million health workers are needed, and this shortage of medical doctors and other health workers trained to deliver HIV treatment and care has been identified as the most serious barrier to the sustained scale up of ART in resource-poor settings.
Task-shifting involves the delegation of healthcare tasks from more highly trained individuals to those with less training, and it has been increasingly employed as a way to help address the shortage of highly trained staff in many resource-limited settings.
Few countries in the developing world legally mandate non-physician clinicians to provide HIV treatment and care (exceptions include Malawi, Kenya, Ethiopia and Uganda). There is a reluctance to legalise task-shifting due to concerns about the capacity of the workers, as well as the quality of care provided.
Conversely this is a common and very successful practice in the developed world where non-physician clinicians (in particular nurse practitioners) play a central role in the treatment and care of HIV-positive patients.
Moreover the Integrated Management of Adolescent and Illnesses (IMAI) guidance, developed by the World Health Organization and partners, uses task-shifting to support a public health approach to treating HIV within government health systems. The target audience is principally healthcare workers providing clinical care at front-line health facilities. More than 30 countries, mostly in sub-Saharan Africa are currently using IMAI.
Ashwin Vasan and colleagues designed a study in which a non-physician first assessed the patient and made a recommendation on whether to start treatment, after which a physician repeated the procedure (without being aware of the initial recommendation). The study compares each worker's assessment, although all final clinical decisions were made by the physician.
Study sites were twelve district hospitals and subdistrict primary care clinics in Uganda. Sites were chosen based on their large HIV programmes, and staff were trained in the management of HIV through the Ministry of Health programme (an adaptation of generic WHO/IMAI protocols). Rural or semi rural sites were chosen to provide an accurate picture of decentralised ART care. Viral load testing was not available and access to CD4 cell counts was limited at all sites.
Healthcare workers were classified as:
- Physicians: those who had completed a six year medical school programme plus a one-year internship. Physicians were not usually HIV specialists.
- Clinical officers: those who had completed three years' pre-service education plus two years' internship.
- Nurses, who had completed between one and four years' formal nursing education.
The authors acknowledge that the variation in the level of training of Ugandan nurses is a limitation of the study. Moreover, the study did not record the number of years of work experience that each worker had.
The researchers used Kappa analysis to evaluate levels of agreement. Kappa is a statistical measurement of the degree of agreement between two observers, and which rates strength of agreement as ‘slight’ (0 to 0.2), ‘fair’ (0.2 to 0.4), ‘moderate’ (0.4 to 0.6), ‘substantial’ (0.6 to 0.8) or ‘almost perfect’ (0.8 to 1).
The 521 eligible patients were HIV-positive adults not currently on treatment.
Each patient was first assessed by either a clinical officer or a nurse, and then by a physician. The primary outcome in the study was the final recommendation on whether or not to start antiretroviral therapy.
In the clinical officer arm, agreement with physicians on starting therapy was almost perfect (Kappa 0.91), with 95% of decisions being the same (as opposed to 43% agreement which could be expected by chance alone).
Agreement between nurses and physicians was less consistent and was calculated to be at the top end of the moderate category (Kappa 0.59). A total of 78% of decisions were in agreement, as opposed to 46% which could be expected by chance alone.
Moreover, both nurses and clinical officers were both somewhat more likely than physicians to recommend an alternative to the standard d4T/3TC/nevirapine regime.
Nurses’ assessments of patients’ WHO clinical stage and TB status were both in substantial agreement with those of physicians. On the other hand, the assessments of clinical officers for both these items were only in moderate agreement with the doctors.
On a number of items, the agreement of both nurses and clinical officers with physicians was assessed as ‘fair’. This was the case for functional status (able to work, able to walk, etc.), opportunistic infection status, absolute exceptions to starting antiretroviral therapy immediately, and patient readiness to start therapy. However the authors note that a number of these variables are subjective in nature. Moreover in some cases there was a great deal of missing data.
Although nurses had higher agreement scores on some items than clinical officers, the authors focus their conclusions on the latter group because of the decisions concerning initiation of therapy. They say that there is “compelling evidence” that clinical officers should be allowed to initiate therapy.
The authors believe that their findings show that under routine conditions at rural health facilities, and without any targeted increase in training or supervision of healthcare workers beyond the national framework, there is agreement in the clinical judgement between different cadres of healthcare workers in terms of starting antiretroviral therapy.
Nonetheless, they also describe the study as a pilot, and argue that “these preliminary data warrant more detailed and multicountry investigation”.