Abstract:The objective of this study was to identify the relationship between hospitalization treatment strategies leading to change in symptoms during 12-week follow-up among hospitalized patients during the COVID-19 outbreak. In this article, data from a prospective cohort study on COVID-19 patients admitted to Khorshid Hospital, Isfahan, Iran, from February 2020 to February 2021, were analyzed and reported. Patient characteristics, including socio-demographics, comorbidities, signs and symptoms, and treatments during hospitalization, were investigated. Also, to investigate the treatment effects adjusted by other confounding factors that lead to symptom change during follow-up, the binary classification trees, generalized linear mixed model, machine learning, and joint generalized estimating equation methods were applied. This research scrutinized the effects of various medications on COVID-19 patients in a prospective hospital-based cohort study, and found that heparin, methylprednisolone, ceftriaxone, and hydroxychloroquine were the most frequently prescribed medications. The results indicate that of patients under 65 years of age, 76% had a cough at the time of admission, while of patients with Cr levels of 1.1 or more, 80% had not lost weight at the time of admission. The results of fitted models showed that, during the follow-up, women are more likely to have shortness of breath (OR = 1.25; P-value: 0.039), fatigue (OR = 1.31; P-value: 0.013) and cough (OR = 1.29; P-value: 0.019) compared to men. Additionally, patients with symptoms of chest pain, fatigue and decreased appetite during admission are at a higher risk of experiencing fatigue during follow-up. Each day increase in the duration of ceftriaxone multiplies the odds of shortness of breath by 1.15 (P-value: 0.012). With each passing week, the odds of losing weight increase by 1.41 (P-value: 0.038), while the odds of shortness of breath and cough decrease by 0.84 (P-value: 0.005) and 0.56 (P-value: 0.000), respectively. In addition, each day increase in the duration of meropenem or methylprednisolone decreased the odds of weight loss at follow-up by 0.88 (P-value: 0.026) and 0.91 (P-value: 0.023), respectively (among those who took these medications). Identified prognostic factors can help clinicians and policymakers adapt management strategies for patients in any pandemic like COVID-19, which ultimately leads to better hospital decision-making and improved patient quality of life outcomes.

Association of the medication protocols and longitudinal change of COVID-19 symptoms: a hospital-based mixed-statistical methods study