Feasibility of screening for cognitive impairment among older persons and referral by community health workers in Wakiso district, Uganda

BACKGROUND: In Uganda, cognitive impairment in older persons aged ≥ 60 years is often undiagnosed due to inadequate appreciation of the condition compounded with limitations of trained human resource able to conduct appropriate cognitive evaluations. Use of Community Health Workers (CHWs) especially in hard-to-reach communities can be an important link for older persons to the health facilities where they can receive adequate evaluations and interventions for cognitive challenges. The aim of the study was to assess the feasibility of screening for cognitive impairment among older persons and referral by CHWs in Wakiso district, Uganda. METHODS: This was a sequential explanatory mixed methods study. The CHWs received a one-day training on causes, signs and symptoms, and management of cognitive impairment and screened older persons ≥ 60 years for cognitive impairment using the Alzheimer’s Disease scale 8 (AD8). Psychiatric clinical officers (PCOs) administered the AD8 and the Mini Mental State Examination to the older persons after assessment by the CHWs who then referred them for appropriate clinical care. We conducted Kappa statistic for agreement between the CHWs and PCOs and compared raw scores of the CHWs to Experts scores using Bland Altman and pair plots and corresponding analyses. We also conducted focus group discussions for the older persons, caregivers and CHWs. RESULTS: We collected data from 385 older persons. We involved 12 CHWs and 75% were females, majority were married (58.3%) with at least a secondary education (66.7%). There was 96.4% (CI 94.5–98.2%) agreement between PCOs and CHWs in identifying cognitive impairment with the PCOs identifying 54/385 (14.0: 95%CI 10.7–17.9%) older persons compared to 58/385 (15.1: 95%CI 11.6–19.0%) identified by CHWs. Of the 58 identified to have cognitive impairment by the CHWs, 93.1% were referred for care. The average difference between the score of the expert and that of the CHW was − 0.042 with a 95% CI of -1.335 to 1.252. Corresponding Bland Altman and pair plots showed high agreement between the measurements although CHWs scored higher values with increasing scores. CONCLUSION: CHWs can be trained to identify and refer older persons with cognitive impairment in the communities. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12888-023-05015-0.