Determinants of Hospital Length of Stay among Burn Patients Using Quantile Regression

Abstract

Background and Objectives Burns is one of the most common and potentially dangerous public health problems. Burn patients' hospitalization facilitates the provision of medical services. However, prolonging the length of hospital stay can not only impose an economic cost, but also cause various infections in patients. Various factors affect patients’ hospitalization length including patients age and gender, burn cause, anatomic place and its severity. Identifying and considering these factors will reduce burn patient’s hospitalization and its consequences. The aim of this study was to evaluate the factors affecting hospitalization length in burn patients admitted to Sina Hospital as burn referral hospital in Northwest of Iran. Material and Methods In this cross-sectional study, the medical records of burn patients who were admitted to Sina Hospital in Tabriz during 2018, were included. Data collection was performed using a goal-based researcher data collection form. The form included patient’s demographics (age, gende, marital status, education level, occupation, comorbidity) and Status of vital signs at admission including respiratory rate, heart rate, body temperature, Glasgow Coma Scale score (less than 12 moderate injuries, 13 and 14 mild injuries, 15 or more normal) and systolic and diastolic blood pressure, burn place, date, cause, TBSA, burn severity, length of stay and outcome. Data were analyzed using Stata 16 software and through quantile regression modeling. Data normality was evaluated using Kolmogorov-Smirnov test, skewness and kurtosis. Due to the skewness of the hospitalization variable distribution (hospitalization length was a non-negative variable with right skew and skewness coefficient of 4.72 and kurtosis coefficient equal to 6.53), to obtain a complete picture of how the conditioned distribution of the response variable changes from variables Independently, Quantile regression was used for modeling with a significance level of P <0.05. Results The total number of hospitalized burn patients was 1586 of whom 998 (62.3%) were male. The mean age of the patients was 25.5 ± 22.9 years. Most of the patients were in age-group 0-5 years. About 22.3% of patients were in Tachypnea situation and 33.3% had low blood pressure. Based on Glasgow score, 0.3% of patients were with moderate and 2.2% with low injury. Only 15.4% of patients were with more than 20% burn in their body. Burn degree III was the most prevalent injury (35.1%). Upper limbs (62.9%) were the most anatomic sections injured in burn. Totally 94 patients (5.9%) were died due to burn injuries. Burns occurred more often in summer (30.5%) and at home (n=1246, 78.6%). The median length of hospitalization was 8 days (95% CI: 7.34-8.57). Single (P = 0.010) and illiterate patients (P = 0.022) had a longer hospitalization length. The lower the Glasgow coma score at admission, resulted in longer hospital stay (P = 0.034). Patients with burn on face (P = 0.037), head and neck (P <0.001) and back of the trunk (P = 0.031) had longer hospital stay, respectively. The higher the percentage of burns, the longer the hospital stay (P <0.001) and the presence of concurrent trauma also increased the length of hospital stay (P <0.001). Conclusion Glasgow coma score, burns on the face, neck and back were identified as effective clinical signs on hospitalization length of stay amongst burn patients. Considering these factors in the triage of burn patients, providing quality treatment and care services to manage these symptoms can reduce the length of hospital stay and ultimately lead to a reduction in social and economic costs for patients and society. Extended Abstract Background and Objectives Burns is one of the most common and potentially dangerous public health problems. Burn patients' hospitalization facilitates the provision of medical services. However, prolonging the length of hospital stay can not only impose an economic cost, but also cause various infections in patients. Various factors affect patients’ hospitalization length including patients age and gender, burn cause, anatomic place and its severity. Identifying and considering these factors will reduce burn patient’s hospitalization and its consequences. The aim of this study was to evaluate the factors affecting hospitalization length in burn patients admitted to Sina Hospital as burn referral hospital in Northwest of Iran. Material and Methods In this cross-sectional study, the medical records of burn patients who were admitted to Sina Hospital in Tabriz during 2018, were included. Data collection was performed using a goal-based researcher data collection form. The form included patient’s demographics (age, gende, marital status, education level, occupation, comorbidity) and Status of vital signs at admission including respiratory rate, heart rate, body temperature, Glasgow Coma Scale score (less than 12 moderate injuries, 13 and 14 mild injuries, 15 or more normal) and systolic and diastolic blood pressure, burn place, date, cause, TBSA, burn severity, length of stay and outcome. Data were analyzed using Stata 16 software and through quantile regression modeling. Data normality was evaluated using Kolmogorov-Smirnov test, skewness and kurtosis. Due to the skewness of the hospitalization variable distribution (hospitalization length was a non-negative variable with right skew and skewness coefficient of 4.72 and kurtosis coefficient equal to 6.53), to obtain a complete picture of how the conditioned distribution of the response variable changes from variables Independently, Quantile regression was used for modeling with a significance level of P <0.05. Results The total number of hospitalized burn patients was 1586 of whom 998 (62.3%) were male. The mean age of the patients was 25.5 ± 22.9 years. Most of the patients were in age-group 0-5 years. About 22.3% of patients were in Tachypnea situation and 33.3% had low blood pressure.Based on Glasgow score, 0.3% of patients were with moderate and 2.2% with low injury. Only 15.4% of patients were with more than 20% burn in their body. Burn degree III was the most prevalent injury (35.1%). Upper limbs (62.9%) were the most anatomic sections injured in burn. Totally 94 patients (5.9%) were died due to burn injuries.Burns occurred more often in summer (30.5%) and at home (n=1246, 78.6%). The median length of hospitalization was 8 days (95% CI: 7.34-8.57). Single (P = 0.010) and illiterate patients (P = 0.022) had a longer hospitalization length. The lower the Glasgow coma score at admission, resulted in longer hospital stay (P = 0.034). Patients with burn on face (P = 0.037), head and neck (P <0.001) and back of the trunk (P = 0.031) had longer hospital stay, respectively. The higher the percentage of burns, the longer the hospital stay (P <0.001) and the presence of concurrent trauma also increased the length of hospital stay (P <0.001). Conclusion Glasgow coma score, burns on the face, neck and back were identified as effective clinical signs on hospitalization length of stay amongst burn patients. Considering these factors in the triage of burn patients, providing quality treatment and care services to manage these symptoms can reduce the length of hospital stay and ultimately lead to a reduction in social and economic costs for patients and society. Practical implications of research Paying attention to the effective symptoms during the stay of burn patients during triage of these patients and providing quality treatment and care services to manage these symptoms can reduce the length of hospital stay and ultimately lead to a reduction in its social and economic costs. For patients and the community. Ethical considerations This research was carried out after the approval of the Ethics Committee of the Vice Chancellor for Research of Tabriz Azad University of Medical Sciences, as a dissertation for general medicine. Data related to patients with burn injuries admitted to Sina Hospital in 1397 were coded in the form of data registration without name and information. Also, all patient information was kept completely confidential and the information obtained from the study was used for research purposes only. Conflict of interest The authors state that there is no conflict of interest in publishing this article. Acknowledgement The present article is taken from the dissertation of General Doctor of Medicine. The authors consider it necessary to express their gratitude to the Vice Chancellor for Research and Technology of Tabriz Azad University of Medical Sciences and the director of Sina Hospital and the people participating in the research.