Length of stay after colorectal surgery in Italy: the gap between “fit for” and “actual” discharge in a prospective cohort of 4529 cases

Abstract

Background

It is common to observe a gap between the day on which the discharge criteria are reached and the actual day of discharge after colorectal surgery. The aim of this study is to understand the reasons for this difference and its clinical impact on the overall length of stay (LOS).

Methods

All patients enrolled in the prospective iCral3 study were analyzed regarding any difference and reason between the “fit for discharge” (FFD) and “actual discharge” (AD) dates. The association between the gap and the LOS in the whole population was then assessed through a multivariate regression model including other confounding variables.

Results

The analysis included 4529 patients, with a median [IQR] LOS of 6 [4–8] days. The median [IQR] LOS was 6 [4–8] days in the no-gap group (3,910 patients, 86.3%), significantly lower (p < .001) than 7 [6–10] days in the gap group (619 patients, 13.7%). Among the gap reasons, the “need for postoperative rehabilitation” compared to “not willing to return home” and “social constraints” was associated with the longest LOS (9 [6.0–12.5] days, p < 0.001 vs other reasons). The existence of the gap independently determined a 2.3-day lengthening of LOS.

Conclusions

Among other factors, the gap between FFD and AD had an independent impact on LOS. The most frequent reasons for this gap were “not willing to return home” and “social constraint”, while the “need for postoperative rehabilitation” had the greater clinical impact.

Introduction

In recent years, enhanced recovery after surgery (ERAS) pathways have been developed to improve perioperative care (Gustafsson et al. 2019). Briefly, ERAS is a multidisciplinary, patient-centered protocol in which clinical standards are implemented using evidence-based practice. A clear effect of high adherence to the ERAS pathway in reducing adverse events and shortening the overall length of stay (LOS) has been demonstrated in colorectal surgery (CRS) (Catarci et al. 2023) and in other surgical specialties (Mithany et al. 2023). LOS is one of the outcome measures, considered a proxy of in-hospital recovery because hospital discharge assumes that a minimal level of physiological and functional recovery has already been achieved (Balvardi et al. 2018).

Although current guidelines (Gustafsson et al. 2019; Ficari et al. 2019) underline the importance of fulfilling several criteria for home discharge, LOS after CRS remains highly variable (Balvardi et al. 2018). Furthermore, the definition of “fit for discharge” (FFD) remains non-homogeneous in the pertinent literature (Maessen et al. 2008), and in clinical practice, it may not be sufficient to define adequate ad hoc criteria. Therefore, actual discharge (AD) can be postponed for some causes related to local culture, surgeons’ preferences, patient expectations, health systems, and insurance status (Balvardi et al. 2018). The present study aimed to measure and analyze the reasons for the gap between FFD and AD and its impact on LOS in patients enrolled in a prospective multicenter observational study of the Italian ColoRectal Anastomotic Leakage study group (iCral3 study), designed to investigate the effect of adherence to the ERAS pathway on patient-reported outcome measures and return to intended oncologic therapy after CRS (Italian ColoRectal Anastomotic Leakage (iCral3) study group 2023).

Methods

This was a planned secondary analysis of a prospective multicenter study conducted in Italy from October 2020 to September 2021 in 76 surgical centers participating in the iCral3 study on a voluntary basis (Italian ColoRectal Anastomotic Leakage (iCral3) study group 2023). All patients who underwent elective or delayed urgency (≥ 24 h from admission) CRS with anastomosis were assessed for inclusion according to explicit inclusion and exclusion criteria. The inclusion criteria were as follows: (a) patients who underwent laparoscopic/robotic/open/converted colorectal resection with anastomosis, including planned Hartmann’s reversals; (b) American Society of Anesthesiologists’ (ASA) class I, II, or III; c) elective or delayed urgency (≥ 24 h from admission) surgery; and d) patients’ written acceptance to be included in the study. The exclusion criteria were as follows: (a) ASA class IV-V, (b) emergent (≤ 24 h from admission) surgery, (c) pregnancy, and (d) hyperthermic intraperitoneal chemotherapy for carcinomatosis. Finally, 4529 patients were enrolled (Italian ColoRectal Anastomotic Leakage (iCral3) study group 2023). Both AD and FFD were prospectively recorded for every enrolled patient by local investigators. Fitness for discharge was defined using international consensus criteria (Fiore et al. 2012): independent mobilization, adequate oral intake, stool passage, good pain control with oral analgesics, no evidence of complications, and patient consent to return home. Each participating center was defined as high volume (> 4 cases) or low volume (≤ 4 cases) according to the median number of enrolled cases per month. The existence of an institutional enhanced recovery pathway (ERP) (having a locally implemented ERP team and protocol, supported by a specific resolution of the hospital/company strategic management) was declared by 48 out of 76 (63.1%) participating centers. The ERAS pathway was defined on 26 items derived from international guidelines (Gustafsson et al. 2019), with percentage adherence measured through explicit criteria (Supplemental Table S1) and reported as below or above the median and below/at or above the 4th quartile. Surgical procedures were categorized as standard (anterior resection, right colectomy, and left colectomy) versus non-standard (splenic flexure resection, transverse colectomy, Hartmann’s reversal, subtotal and total colectomy, and other) resections. Anastomotic leakage (AL) was defined and graded according to international consensus (Rahbari et al. 2010). Nutritional status, measured using the Mini Nutritional Assessment–Short Form (MNA-SF) (Kaiser et al. 2009), was categorized as below or above 12, indicating the threshold of normal nutritional status.

Outcomes

All outcomes were calculated 60 days after surgery. Any adverse events were recorded, classified, and graded according to the Clavien-Dindo (Dindo et al. 2004) and the Japanese Clinical Oncology Group extended criteria (JCOG-EC) (Katayama et al. 2016), as well as any unplanned readmission, reoperation, and death. Overall morbidity (OM) was defined as any adverse event, while major morbidity (MM) was defined as any major adverse event (Clavien-Dindo grade > II). Thereafter, the study population was divided into two subgroups: (a) patients with no gap between FFD and AD (no-gap group); and (b) patients in which this gap was present (gap group). The primary endpoint was the effect of the gap on LOS, calculated by including any planned or unplanned readmission.

Data collection and statistical analysis

All data of the enrolled patients were prospectively recorded in a web-based database via an electronic case report form, specifically designed for the study and protected by access credentials for each center/investigator. Continuous and discrete variables related to biometric data, patient-related risk factors, indication and type of surgical procedure, adherence to ERAS pathway items, and outcomes were recorded (Supplemental Table S2). Quality control of the data for consistency, plausibility, and completeness was performed on every single record by local investigators and subsequently validated by the study coordinator, resolving any discrepancies through strict cooperation. During the perioperative period, patients were examined daily by local investigators, who were free to decide on complementary imaging and any further action according to the local criteria. Data were reported as medians (IQR) for continuous variables and numbers (percentages) for categorical variables. As most variables were not normally distributed by the Kolmogorov–Smirnov test, between-group comparisons were performed using the Mann–Whitney U test and the Kruskal–Wallis test for continuous variables, and the Pearson chi-square test or the Fisher exact test was used for categorical variables. Univariate and multivariate regression analyses adjusted for confounders were used to evaluate independent effects on LOS. All variables with univariate P < 0.10 were tested in a multivariate regression model performed using the Enter procedure. Standardized beta coefficients and model-corrected R² values were also determined. Collinearity diagnostics were performed to check model stability, by assessing tolerances, variance inflation factors, and condition indexes. For each variable identified as significantly associated with LOS in the multivariate analysis, we also reported the difference in LOS between the two groups identified by that variable. Statistical significance was set at P value < 0.05. All analyses were performed using SPSS (Statistical Package for Social Sciences) for Windows (version 22.0; SPSS Inc., Chicago, IL, USA).

Ethics

The study was conducted in accordance with the Declaration of Helsinki and the principles of the guidelines for good clinical practice E6 (R2). The study protocol was registered at ClinicalTrials.gov (NCT04397627) after approval by the ethics committee of the coordinating center (Marche Regional Ethics Committee-CERM 2020/192 released on 07/30/2020) and the local ethics committee of all other participating centers. Individual participant-level anonymized datasets were made available upon explicit request by contacting the study coordinator.

Results

A total of 4,529 patients were included in this study, of which 68% received preoperative ERAS education (Supplemental Tables S1 and S2). The patients were followed for a median of 68 days (IQR 59–112). The baseline characteristics of the entire study population and subgroups are shown in Table 1. A gap between FFD and AD was recorded in 619 (13.7%) patients (gap group), and there was no gap in the remaining 3910 (86.3%) patients (no-gap group). The median (IQR; range) LOS in the entire population was 6.0 (4.0–8.0; 2–91) days; it was 6.0 (4.0–8.0; 2–91) days in the no-gap group and 7.0 (6.0–10.0; 3–79) days in the gap group (P < 0.001). Gap reasons were specified in 605 (97.7%) patients, the most frequent being “not willing to return home” (229 patients; 37.8%) and “social constraint” (240 patients; 39.7%). A “need for postoperative rehabilitation” was reported in 136 (22.5%) patients. It was associated with a significantly longer median [IQR] LOS of 9.0 [6.0–12.5] days, compared to the other two reasons (7.0 [6.0–9.0] days for both, P < 0.001) (Fig. 1). Twenty-one of 28 variables considered in the study had a significant influence on LOS in univariate analysis and were included in the multivariate regression model (Table 2). An independent increase in LOS was recorded for patients aged > 69 years and ASA class > II among patient-related variables; neo-adjuvant therapy, intra-postoperative blood transfusions, delayed urgency admission, non-standard procedure, procedure length > 180 min, associated procedures, and discharge gap among treatment-related variables; anastomotic leakage, overall morbidity, major morbidity, and reoperation among outcome-related variables. In contrast, both minimally invasive surgery and ERAS adherence > 69.2% were independently associated with a reduction of LOS. The quantitative differences with baseline LOS related to these variables are shown in Fig. 2.

Table 1 Baseline characteristics of the study population and subgroups

Fig. 1

Comparison of LOS based on gap causes within the gap group

Table 2 Influence of patient-, treatment-, and outcome-related factors on the length of stay: univariate and multivariate regression analyses

Fig. 2

Quantitative differences with baseline related to the independent predictors of LOS

Discussion

This observational study revealed the existence of a gap between FFD and AD after CRS in 13.7% of cases. This figure is lower than those reported in recent retrospective (Slieker et al. 2017) and prospective analyses (Biondi et al. 2022), where the concordance between FFD and AD barely reached one-third of the cases. This result can be attributed to several factors. In recent years, many improvements in the enhanced recovery pathway have been implemented: the return to home will probably be more anticipated by patients, they would be better prepared, and it would be easier for the treating physician to make an announcement about the discharge; surgeon-related perception of potential risks associated with early discharge that could lead to readmission and/or continued hospitalization is mitigated by the large body of evidence that makes discharge safe (Jones et al. 2017). In the prospective multicenter iCral1 study (Italian ColoRectal Anastomotic Leakage (iCral) Study Group. Anastomotic leakage after elective colorectal surgery: a prospective multicentre observational study on use of the Dutch leakage score, serum procalcitonin and serum C-reactive protein for diagnosis 2020), the use of clinical scores with serum biomarkers was demonstrated to have a high (< 99%) negative predictive value for anastomotic leakage already on postoperative day 2. On the other hand, a prospective French study designed to evaluate the feasibility and safety of a 3-day hospitalization after colectomy and 5-day hospitalization after proctectomy in patients enrolled in an enhanced recovery program showed that serum biomarkers above the threshold accounted for delayed discharge after CRS in 15% of cases, while patient refusal was the most common reason (Collard et al. 2020), as it was in the present analysis. Patient and caregiver education, contact information, and discharge planning are increasingly recognized as a fundamental part of surgical patient care (Biondi et al. 2022; Jones et al. 2017; Chugh et al. 2009), and improving the rate of adherence to preoperative counseling (limited to 68% of cases in the present study, Supplemental Table S2) may limit patients’ refusal to return home when the criteria for safe discharge are met.

This study also reported that a median ERAS adherence of 69.2% determined a median LOS of 6 days, which is equal to the constant value reported by a nationwide study in the USA on 251,583 patients who underwent elective CRS from 2002 to 2011 (Al-Mazrou et al. 2017). This figure corresponds to a reduction of 1 day compared with the Italian nationwide LOS after laparoscopic CRS recorded in the same period (Agenzia Nazionale per i Servizi Sanitari Regionali (AGENAS) 2023). While the reduction in LOS recorded in randomized trials on ERAS generally exceeds 2 days (Greco et al. 2014), the results of the present study are not different from those achieved in population-based studies. A pre-post structured ERAS implementation study among 15 teaching hospitals in Ontario, Canada, performed on 32,612 patients who underwent CRS between 2008 and 2019 (Bayat et al. 2024), revealed a mean (95%CI) reduction of LOS at 1.05 (0.72–1.38) days. This finding confirms that complete implementation of any ERP is not an “on–off” process, as it involves deep cultural and behavioral changes in patients and healthcare providers and requires time and perseverance to reach a level of implementation (measurable through the ERP adherence rate) that can influence LOS and other outcomes. The existence of a self-declared institutional ERP by 48 out of 76 (63.1%) centers participating in the present study had no influence on LOS (Table 2), confirming that "declaring an ERAS protocol is not enough” (Maessen et al. 2007). Structured implementation and auditing processes are more important for improving program adherence and outcomes (Catarci et al. 2021).

Comparison of the no-gap and gap groups (Table 1) disclosed that the latter contains significantly more “bad performers”, such as older and comorbid patients, higher rate of delayed urgency presentation, and more complex operations (longer operative time with higher rates of blood transfusions and associated procedures). This could explain the “social constraint” and “need for postoperative rehabilitation” as the reasons for the gap in two-thirds of the cases in the present study. CRS is increasingly performed in older patients with multiple comorbidities, and surgical procedures are often more complex; these factors may intuitively challenge any smooth recovery process. In this setting, anticipating the need for postoperative intensive care (Conti et al. 2023), adding a case manager nurse dedicated to ERAS programs, and implementing referral pathways for postoperative rehabilitation should be provided to improve postoperative compliance (Li et al. 2013).

Finally, the multivariate analysis for LOS in the entire cohort (Table 2) showed that any adverse event (particularly major events such as anastomotic leakage) and blood transfusions had the highest weight on the independent lengthening of LOS, as expected (Italian ColoRectal Anastomotic Leakage (iCral) study group 2019; Catarci et al. 2022; Italian ColoRectal Anastomotic Leakage (iCral) study group 2020; Catarci et al. 2023). While it appears straightforward that patient-related factors, such as older age and multiple comorbidities, and treatment-related factors, such as delayed urgency presentation, longer and associated procedures, neo-adjuvant therapy, and non-standard resections, also had an independent effect on LOS, the discharge gap independently determined a 2.3-day lengthening of LOS (Fig. 2).

The study data were obtained from a large multicenter prospective investigation of patients undergoing colorectal surgery, which was performed at a well-defined time-lapse in a large number of centers representing a wide sample of surgical units performing colorectal surgery in Italy. However, this study has some limitations. The nature of this study does not permit a definitive conclusion about this topic. Another limitation is the potential for residual, measured, and unmeasured confounding intrinsic to observational studies. Although data quality control was performed and repeated at various levels, we could not exclude any measurement errors from the participating investigators. Finally, the study results were obtained using a non-a priori sample structured for this topic.

In conclusion, although more studies focused on this topic are needed, our analysis pointed “not willing to return home” and “social constraint” as the most frequent reasons for the gap between FFD and AD, while “needing rehab” had the greater clinical impact. This gap, along with other variables, independently impacts LOS lengthening after CRS.