Identification of patients eligible for discharge within 48 h of colorectal resection.

ABSTRACT

BACKGROUND: This study aimed to identify patients eligible for a 48-h stay after colorectal resection, to provide guidance for early discharge planning. METHODS: A bi-institutional retrospective cohort study was undertaken of consecutive patients undergoing major elective colorectal resection for benign or malignant pathology within a comprehensive enhanced recovery pathway between 2011 and 2017. Overall and severe (Clavien-Dindo grade IIIb or above) postoperative complication and readmission rates were compared between patients who were discharged within 48 h and those who had hospital stay of 48 h or more. Multinominal logistic regression analysis was performed to ascertain significant factors associated with a short hospital stay (less than 48 h). RESULTS: In total, 686 of 5122 patients (13·4 per cent) were discharged within 48 h. Independent factors favouring a short hospital stay were age below 60 years (odds ratio (OR) 1·34; P = 0·002), ASA grade less than III (OR 1·42; P = 0·003), restrictive fluid management (less than 3000 ml on day of surgery OR 1·46; P < 0·001), duration of surgery less than 180 min (OR 1·89; P < 0·001), minimally invasive approach (OR 1·92; P < 0·001) and wound contamination grade below III (OR 4·50; P < 0·001), whereas cancer diagnosis (OR 0·55; P < 0·001) and malnutrition (BMI below 18 kg/m2 OR 0·42; P = 0·008) decreased the likelihood of early discharge. Patients with a 48-h stay had fewer overall (10·8 per cent versus 30·6 per cent in those with a longer stay; P < 0·001) and fewer severe (2·6 versus 10·2 per cent respectively; P < 0·001) complications, and a lower readmission rate (9·0 versus 11·8 per cent; P = 0·035). CONCLUSION: Early discharge of selected patients is safe and does not increase postoperative morbidity or readmission rates. In these patients, outpatient colorectal surgery should be feasible on a large scale with logistical optimization.

RESUMEN

ANTECEDENTES Este estudio tuvo como objetivo identificar pacientes candidatos para una estancia hospitalaria de 48 horas tras resecciones colónicas, con el fin de proporcionar una guía de planificación del alta precoz. MÉTODOS: Estudio de cohortes retrospectivo de pacientes consecutivos sometidos a resección colorrectal electiva mayor por patología benigna o maligna en el marco de un programa integral de recuperación intensificada (enhanced recovery pathway, ERP), de dos hospitales entre 2011 y 2017. Se compararon las tasas de complicaciones postoperatorias globales y graves (Clavien ≥ IIIb) y de reingresos entre dos grupos (< 48 horas versus ≥ 48 horas de estancia hospitalaria). Se llevó a cabo una regresión logística multinominal de factores significativos (P < 0,05) asociados con una estancia corta (< 48 horas). RESULTADOS: En total, 686/5.122 pacientes (13,4%) fueron dados de alta dentro de las primeras 48 horas. Los factores independientes que propiciaron una estancia corta fueron la edad < 60 años (razón de oportunidades, odds ratio, OR 1,34, P = 0,002), puntuación < 3 de la American Society of Anesthesiologists (ASA) (OR 1,42, P = 0,003), manejo restrictivo del aporte de líquidos (< 3000 mL en el día de la cirugía OR 1,46, P < 0,001), duración de la cirugía < 180 minutos (OR 1,89, P < 0,001), abordaje mínimamente invasivo (OR 1,92, P < 0,001) and tipo de herida clase < 3 (OR 4,5, P < 0,001), mientras que el diagnóstico de cáncer (OR 0,55, P < 0,001) y la malnutrición (IMC < 18 kg/m2 OR 0,42, P = 0,008) disminuyeron la probabilidad de alta precoz. Los pacientes con una estancia de 48 horas tuvieron menos complicaciones globales (10,8% versus 30,6%, P < 0,001), menos complicaciones graves (2,6% versus 10,2%, P < 0,001) y una menor tasa de reingresos (9% versus 11,8%, P = 0,035). CONCLUSIÓN: El alta precoz en pacientes seleccionados es segura y no aumenta las tasas de morbilidad postoperatoria o de reingresos. En estos pacientes, la cirugía colorrectal ambulatoria debería ser viable a gran escala con una optimización de la logística.