Factors associated with hospitalisation among diabetes patients in Western Kenya

Abstract

BACKGROUND

Diabetes patients routinely visit hospitals and often get admitted for periods. However, drivers of routine hospital visits and admissions (here defined as hospitalisation) are poorly understood to inform improved management of diabetes at the household level. This study determined factors associated with hospitalisation among diabetes patients in Busia County, Kenya.

METHODOLOGY

The study adopted a hospital-based cross-sectional study design and employed systematic sampling to select 155 participants from diabetes patients seeking healthcare services at Busia County Referral Hospital. Data was collected in March 2023, six months retrospectively using a structured questionnaire. Bivariate and adjusted multivariable logistic regressions were used to assess factors associated with hospitalisation.

RESULTS

More than one-half (n = 89, 57.4%) of the participants were females. Over 85% of the participants enrolled were aged 40 years and above (40-59 years [n = 66, 42.6%]; 60+ years [n = 67, 43.2%]). Approximately 53% of respondents reported at least one hospital admission within six months, whereas 47% had routine visits. Participants aged at least 60 years old had higher odds of hospital admission and routine clinic visits; 16.07 (Adjusted Odds Ratio [AOR] = 16.07, 95% CI = 4.47 – 57.72, p-value <0.001) and 7.30 (AOR = 7.30, 95% CI = 2.35 – 22.62, p-value <0.010), respectively. Diabetes patients who reported to be taking alcohol had a higher proportion of hospital admissions, p-value = 0.004. Diabetes patients with concurrent HIV had more routine hospital visits compared to their counterparts without HIV (AOR =2.74, 95% CI = 1.03-7.25, p-value <0.001).

CONCLUSIONS

The study shows that old age, employment status, alcohol consumption, and comorbid HIV were associated with hospitalisation. These findings underscore the urgency of targeted interventions to address the socioeconomic disparities and strengthen diabetes management policies in Kenya. Further studies targeting; behavioral status, host diabetes, and comorbidity conditions will be important to inform improved management of diabetes at the household level to reduce hospitalisation.