Quality of Care in Integrated Community Case Management Services in Bugoye, Uganda: a Retrospective Observational Study

Abstract

Background: Village health workers (VHWs) in 5 villages in Bugoye subcounty (Kasese District, Uganda) provide integrated community case management (iCCM) services, in which VHWs evaluate and treat malaria, pneumonia, and diarrhea in children under 5 years of age. VHWs use a “Sick Child Job Aid” that guides them through the evaluation and treatment of these illnesses. Malaria is diagnosed by rapid diagnostic test, pneumonia by age-based respiratory rate cutoffs, and diarrhea by symptoms as described by the caregiver. Objectives: Measure the quality of iCCM services over time using routinely collected iCCM program data. Methods: We used existing aggregate program data to: 1) summarize the patient population and services provided, and 2) measure quality of care based on concordance between matching categories or actions in the treatment algorithm. We then employed lot quality assurance sampling to create a patient-level dataset, and created a secondary patient-level dataset of all patients with “danger signs” (evidence of severe illness). Using patient-level records, we: 1) describe adherence to the iCCM algorithm; 2) measure VHW-level quality using lot quality assurance sampling decision rules; 3) assess change over time in quality of care with generalized estimating equations regression modeling. Results: Most VHWs achieved greater than 90% concordance for all measures apart from concordance between number of patients presenting with fever and number of rapid diagnostic tests performed. From the main patient-level dataset, 97% (150) of patients with diarrhea were treated with oral rehydration and zinc, 95% (216) of patients with presumed pneumonia were treated with amoxicillin, and 94% (240) of patients with malaria were treated with artemisin combination therapy or rectal artesunate. However, only 44% (44) of patients with a negative rapid test for malaria were appropriately referred. Of patients with danger signs, 95% were appropriately referred to a health facility. Overall, 75% (434) of patients received all the correct evaluation and management steps. At the VHW level, 9 out of 23 VHWs have provided high-quality care over 2 years, based on the selected lot quality assurance sampling decision threshold (21 out of 25 patients with correct management). Quality of care increased significantly in the first 6 months after initiation of iCCM services (p = 0.003), and then plateaued in months 7-24. Conclusions: Quality of care was high for uncomplicated malaria, pneumonia and diarrhea. Overall quality of care was lower, partly because VHWs often did not follow the guidelines to refer patients who tested negative for malaria. Quality of care appears to improve as VHWs gain initial experience in iCCM care. Despite some limitations, lot quality assurance sampling and concordance are feasible and scalable approaches to measuring quality of iCCM care.