Evaluation of the effectiveness of implemention of the practical approach to lung health (PALSA) in the Free State

English: Currently, respiratory diseases constitute about one third of patients that present to primary care clinics in under-resourced countries of the world. Communicable respiratory diseases such as tuberculosis, acute respiratory infections in adults and non-communicable respiratory diseases such as asthma, chronic obstructive pulmonary disease, lung cancer represent about one-fifth of the global burden of disease measured in disability adjusted life years (DALY). Opportunistic infections, other respiratory complications, and the widespread use of tobacco further increase the respiratory disease burden in high HIV prevalence settings. In developing countries clinic nurses with limited training and basic skills are entrusted to properly diagnose and treat respiratory patients from overloaded clinics. We developed an educational outreach intervention, Practical Approach to Lung Health in South Africa (PALSA) on integrated respiratory case management aimed at improving the quality of respiratory care in South African primary care clinics. The intervention comprised 3 to 4 academic detailing training sessions of primary care nursing practitioners; dissemination of locally adapted PALSA guidelines and support materials; changes in prescribing provisions for primary care nurses, and doctors' sensitization about PALSA. The impact of PALSA on the processes and outcomes of respiratory care was evaluated through a pragmatic cluster randomized controlled trial in the Free State province in 2003. A total of 1000 patients in the intervention arm and 999 patients in the control arm presenting with respiratory conditions to the 40 largest primary care clinics of the Free State province were interviewed at the first post-intervention survey. The number of patients recruited ranged from 47 to 52 patients per clinic. The follow up rate was 92.9% for the intervention arm and 92.7% for the control arm. Twenty two patients died in the intervention clinics and twenty six died in the control clinics. During data analysis, four patients in each arm were deleted due to unavailability of the first post-intervention survey data and/or because they did not meet the inclusion criteria. Professional nurses in intervention clinics received a median of 2 training sessions while nurses in the control clinic received nothing. First post-intervention survey characteristics of the intervention and control arms balanced as a result of randomization. Almost two thirds of the patients were females with the most frequent age group being 25-54 years. About 50% of patients had a smoking history, about 50% had primary education, close to 50% were unemployed, above 80% walked to get to the nearest clinic and 70% spent between 2 and 12 hours to travel to and from the clinic. The inclusion criteria to the study were adults 15 years and older presenting with a cough or difficulty breathing on the day of the interview, recurrent cough or difficulty breathing in the last 6 months or cough for less than two weeks with any of the four severity markers. Rates of cough and difficulty breathing ranged between 70% and 90%. About 70% of the patients complained about chest symptoms interfering with their usual activities while around 36% had gone to the clinic for a check-up on recurrent respiratory problem. Compared to control clinics, intervention clinics had a significant improvement in inhaled steroid prescription of 16.1% versus 10.3% (odds ratio 1.70; 95%CI 1.13 to 2.56), and an improvement in sending of sputa for tuberculosis testing of 16.7% versus 11.2% (odds ratio 1.60; 95%CI 1.00 to 2.54). There were also significant improvements seen on appropriate referral of patients that had one of the four severity makers of 10.6% versus 4.9% (odds ratio 2.56; 95%CI 1.06 to 6.17), and close to significant improvement of the tuberculosis detection rate of 3.0% versus 1.8% (odds ratio 1.67; 95%CI 0.92 to 3.02). There was a significant increase in interference with usual activities due to chest symptoms of 68.0% versus 60.1% (odds ratio 1.44; 95%CI1.13 to 1.85). There was no improvement on antibiotic prescription of 36.1% versus 38.0% (odds ratio 0.92; 95%CI 0.62 to 1.36) as well as cotrimoxazole prophylaxis of 12.6% versus 9.9% (odds ratio 1.52; 95%CI 0.60 to 3.89). Results of this study suggest that inhaled steroid prescription, tuberculosis case detection rate, and appropriate referral of patients with severe respiratory diseases can be improved in nurse staffed primary care clinics in developing countries and under-resourced settings. This study exemplifies an evaluation of the effectiveness of an educational intervention in South African primary care. It shows how a carefully developed intervention, using a syndromic approach to diagnosis and treatment, can improve several aspects of clinical care after brief training of primary care nurses. It also illustrates opportunities for, and difficulties in, implementing such an intervention, and conducting a large scale trial in this setting. This study suggests that other international interventions based on dissemination of clinical guidelines, such as, for IMCI, STls and HIV/AIDS should be developed and rigorously evaluated locally, given their potential impact on public health and on services.