Clinical complexity and hospital admissions in the December holiday period

Background Christmas and New Year's holidays are risk factors for hospitalization, but the causes of this holiday effect are uncertain. In particular, clinical complexity (CC) has never been assessed in this setting. We therefore sought to determine whether patients admitted to the hospital during the December holiday period had greater CC compared to those admitted during a contiguous non-holiday period. Methods This is a prospective, longitudinal study conducted in an academic ward of internal medicine in 2017-2019. Overall, 227 consecutive adult patients were enrolled, including 106 cases (mean age 79.4 +/- 12.8 years, 55 females; 15 December-15 January) and 121 controls (mean age 74.3 +/- 16.6 years, 56 females; 16 January-16 February). Demographic characteristics, CC, length of stay, and early mortality rate were assessed. Logistic regression analyses for the evaluation of independent correlates of being a holiday case were computed. Results Cases displayed greater CC (17.7 +/- 5.5 vs 15.2 +/- 5.9; p = 0.001), with greater impact of socioeconomic (3.51 +/- 1.7 vs 2.9 +/- 1.7; p = 0.012) and behavioral (2.36 +/- 1.6 vs 1.9 +/- 1.8; p = 0.01) CC components. Cases were also significantly frailer according to the Edmonton Frail Scale (8.0 +/- 2.8 vs 6.4 +/- 3.1; p<0.001), whilst having similar disease burden, as measured by the CIRS comorbidity index. Age (OR 1.02; p = 0.039), low income (OR 1.97, 95% CI 1.10-3.55; p = 0.023), and total CC (OR 1.06; p = 0.014) independently correlated with the cases. Also, cases showed a longer length of stay (median 15.5vs11 days; p = 0.0016) and higher in-hospital (12 vs 4 events; p = 0.021) and 30-day (14 vs 6 events; p = 0.035) mortality. Conclusions Patients hospitalized during the December holiday period had worse health outcomes, and this could be attributable to the grater CC, especially related to socioeconomic (social deprivation, low income) and behavioral factors (inappropriate diet). The evaluation of all CC components could potentially represent a useful tool for a more rational resource allocation over this time of the year.