Impact of abdominal drainage time and bed rest duration on postoperative recovery in patients undergoing laparoscopic appendectomy: a single-center retrospective study

Abstract

Background

The optimal duration of abdominal drainage and bed rest after laparoscopic appendectomy (LA) remains uncertain. This study investigated whether these modifiable factors are associated with residual intra-abdominal infection (RII) and prolonged length of hospital stay (LOS).

Objective

To retrospectively analyze the risk factors for RII and prolonged LOS following LA and clarify the impact of abdominal drainage time and bed rest duration on postoperative recovery.

Methods

This retrospective study included 478 adults who underwent LA for acute appendicitis (2017-2020). Patients were followed weekly for 4 weeks. RII was defined as a composite endpoint occurring between postoperative day 5 and 4 weeks. Multivariable logistic regression identified factors associated with RII and prolonged LOS (≥5 days). Subgroup analyses were exploratory due to small sample sizes.

Results

RII occurred in 59 patients (12.3%). In the overall cohort, independent factors for RII were complicated appendicitis (OR 4.35), drainage ≥36 h (OR 3.69), bed rest ≥24 h (OR 3.27), and Neutrophil-to-lymphocyte ratio (NLR) ≥6.64 (OR 3.18) (all P<0.05). Drainage <36 h was not associated with increased risk. Exploratory subgroup analyses suggested drainage ≥36 h as the only factor for RII in uncomplicated appendicitis (OR 7.47), while bed rest ≥24 h (OR 7.21), elevated White blood cell count (WBC) (OR 4.65), and NLR (OR 6.02) were associated in complicated appendicitis. For prolonged LOS (198 events), independent factors were surgical complications (OR 6.14) and drainage ≥36 h (OR 4.74). Exploratory subgroup analyses showed surgical complications and drainage ≥36 h were significant in uncomplicated appendicitis, while only surgical complications remained in complicated appendicitis.

Conclusion

Prolonged drainage (≥36 h) and bed rest (≥24 h) were associated with increased RII in the overall cohort; however, sensitivity analysis revealed variability in effect estimates, underscoring the need for cautious interpretation. Subgroup findings are exploratory and require confirmation. Prolonged drainage and surgical complications were associated with longer LOS. Drainage <36 h appeared safe. These findings suggest potential benefits of avoiding routine drainage, early drain removal, and early mobilization, but require confirmation in larger prospective studies.

Introduction

The pathogenesis of acute appendicitis (AA) may be related to fecalith obstruction, environmental factors, genetic factors, and infection [1, 2]. Based on pathological type, AA can be classified into simple, suppurative, gangrenous, perforated appendicitis, and periappendiceal abscess. Simple and suppurative appendicitis are termed uncomplicated appendicitis (UA), whereas gangrenous, perforated, and abscess are defined as complicated appendicitis (CA) [1]. UA may be treated conservatively without surgery, but CA is recommended to be treated primarily surgically [1]. The incidence of postoperative complications is significantly higher in CA than in UA, especially infectious complications [3–8], leading to poor postoperative recovery and prolonged length of hospital stay (LOS) [5, 9, 10].

The routine use of prophylactic abdominal drainage after appendectomy remains controversial. Recent meta-analyses have questioned its benefit, suggesting that drainage may increase the risk of intra-abdominal abscess (IAA) and wound infection [11–13]. However, the optimal duration of drainage has rarely been examined.

Early postoperative mobilization is a key component of Enhanced Recovery After Surgery (ERAS) protocols [14, 15] and has been shown to reduce complications and shorten hospital stay after abdominal surgery [16, 17]. Nevertheless, the specific impact of bed rest duration on infectious complications after laparoscopic appendectomy (LA) in adults remains unclear.

Distinguishing clinically significant postoperative infection from normal inflammatory changes is challenging. To avoid overdiagnosis, we applied a 5-day time threshold for residual intra-abdominal infection (RII) diagnosis, based on clinical experience and coinciding with the median hospital stay. This study aimed to identify modifiable factors associated with RII and prolonged length of hospital stay (LOS) after LA, focusing on drainage duration and bed rest time, and to explore whether these associations differ between UA and CA.

Methods

Study design

This retrospective cohort study followed the STROBE guidelines [18]. Data from AA patients who underwent LA between June 30, 2017, and June 30, 2020, were collected. The study was approved by the Ethics Committee of Chongqing Fifth People’s Hospital (2024CQSDWRMYYGS-002) with a waiver of informed consent.

Setting

A retrospective analysis was performed on AA patients treated with LA at Chongqing Fifth People’s Hospital from June 30, 2017, to June 30, 2020. Post-discharge follow-up was conducted weekly by the physician team until 4 weeks after surgery. Data were derived from June 30, 2017, to July 29, 2020.

Participants and follow-up

Inclusion criteria: (1) confirmed AA diagnosis; (2) LA as the primary procedure; (3) age 18–80 years.

Exclusion criteria: (1) primary discharge diagnosis other than AA; (2) conservative treatment > 24 h before surgery; (3) conversion to open surgery; (4) missing follow-up data.

Weekly telephone follow-up was conducted after discharge until 4 weeks postoperatively. Patients with significant discomfort were advised to seek outpatient care. RII diagnosis after discharge was based on outpatient medical records and diagnostic criteria.

Variables and data sources

Treatment

LA was performed via three-port technique. The appendiceal base was ligated with hemoclips or sutured in CA cases. Postoperative care included anti-infection and fluid therapy. Pain was assessed daily using NRS [19]. Analgesia was routinely provided within 24 h, and thereafter for NRS ≥ 4. Oral intake began upon bowel function recovery.

Discharge criteria

Patients were discharged when tolerating semi-liquid diet, with mild pain (NRS < 4) and no significant abdominal signs. Those with elevated inflammatory markers received oral antibiotics for 2–3 days post-discharge.

The decision to place a drain was at the surgeon’s discretion, based on intraoperative findings. Indications for drain placement included: (1) gross peritoneal contamination or localized pus, (2) gangrenous or perforated appendicitis with concern for inadequate source control, (3) incomplete hemostasis, (4) a friable appendiceal stump, and (5) the presence of a periappendiceal abscess. Drains (closed-suction silicone drains) were placed in the pelvic or right paracolic gutter.

Drain removal

Drains were removed when output was < 20 ml/24 h with clear serous fluid.

RII definition

RII was defined as a composite endpoint occurring between postoperative day 5 and 4 weeks, based on postoperative intra-abdominal infection concept [20]. The 5-day minimum was chosen to exclude transient postoperative inflammation. IAA was recorded as a subset of RII. Diagnosis required meeting criterion (1) plus either (2) or (3):

1. Clinical: Abdominal pain (NRS ≥4) with tenderness, with or without fever, rebound tenderness, or guarding (excluding incisional infections).

2. Laboratory: Elevation of at least two inflammatory markers beyond postoperative day 5: WBC > 10.00×10⁹/L, neutrophil percentage >75%, or CRP >10 mg/L.

3. Imaging: Ultrasound or CT evidence of intra-abdominal fluid collection with features suggestive of infection (wall enhancement, internal gas, fat stranding) or abscess formation.

Definition of prolonged bed rest

Bed rest duration ≥ 24 h was defined as having less than 1 h of out-of-bed activity within the first postoperative day.

Data collection

Demographic, preoperative, intraoperative, and postoperative data were collected. CA was defined intraoperatively as gangrenous, perforated, or abscess-forming appendicitis; all others were UA.

Bias

To ensure a homogeneous cohort, cases with a shift in clinical focus away from AA or requiring conversion to alternative surgical approaches were excluded.

Study size and quantitative variables

Of 528 screened patients, 47 were excluded per criteria, and 3 had missing follow-up, leaving 478 for analysis (Fig. 1). Outcome variables were RII and prolonged LOS (≥ 5 days). Continuous variables were categorized using ROC-derived cutoffs.

Fig. 1.

Sample selection process

Statistical methods

Analyses were performed using SPSS 25.0. Group comparisons used t-test, Mann-Whitney U test, or chi-square/Fisher’s exact test as appropriate. Variables with P < 0.05 in univariate analysis were entered into multivariable logistic regression using backward stepwise selection, after checking for multicollinearity (VIF < 2.5). Given the limited RII events (n = 59) relative to candidate variables (10), a sensitivity analysis was performed: the four final variables were forced into a reduced model. Overall LOS model (198 events, 7 variables) had sufficient events and no additional analysis was performed. Subgroup analyses were considered exploratory due to small sample sizes. Statistical significance was set at P < 0.05.

Results

Participants and descriptive data

Of 528 patients who underwent LA and met the inclusion criteria, 47 were excluded based on predefined criteria: 14 had primary discharge diagnoses other than AA, 18 received conservative treatment exceeding 24 h before surgery, and 15 required conversion to open surgery or additional specialized procedures. An additional 3 patients had missing follow-up records. Thus, 478 patients were included in the final analysis (Fig. 1).

Among these, 59 patients (12.3%) were diagnosed with postoperative RII. LOS ranged from 4 to 21 days. Of the 59 RII cases, 10 involved IAA; 9 of these were identified and managed during the initial hospitalization (1 underwent percutaneous catheter drainage, 1 underwent open surgical drainage). The remaining 57 patients with RII recovered with antimicrobial therapy alone. No surgery-related deaths occurred within the 4-week postoperative period.

CA was present in 108 patients (22.6%) and UA in 370 (77.4%). Baseline characteristics are shown in Table 1. Compared to UA patients, those with CA had longer symptom duration, higher temperature, higher inflammatory markers: White blood cell count (WBC), Neutrophil-to-lymphocyte ratio (NLR), C-reactive protein (CRP), longer operative time, more frequent drain placement, longer bed rest, higher Clavien-Dindo classification, and higher rates of postoperative complications, RII, IAA, and prolonged LOS (all P < 0.05).

Table 1.

Baseline and perioperative characteristics by appendicitis severity

Variable	CA(n = 108)		UA(n = 370)		P
Duration(≥ 48 h)	40		88		0.006
Gender(female)	53		170		0.566
Temperature(≥ 37.4 °C)	37		41		<0.001
Preoperative treatment time after admission(≥ 12 h)	6		15		0.593*
Diabetes	7		16		0.178*
Antibiotics(Combined)	13		26		0.094
Retrocecal appendix	13		44		0.967
Adhesiolysis	6		20		0.999
Appendiceal base suture	4		0		0.002*
Appendiceal stone	26		70		0.239
Drainage(no/<36 h/≥36 h)	27/55/26		142/152/76		0.037
Bed rest duration ≥ 24 h	31		65		0.011
Intraperitoneal irrigation	23		21		<0.001
Length of hospital stay(LOS,≥5days)	67		131		<0.001
Wound infection	3		6		0.429*
Ileus	2		3		0.316*
Postoperative bleeding	1		0		0.226*
Intra-abdominal abscess(IAA)	7		3		0.002*
Residual intra-abdominal infection(RII)	38		21		<0.001
Surgical complications(total)	41		29		<0.001
Clavien–Dindo classification(I/II/III/IV/V)	2/38/1/0/0		7/22/1/0/0		<0.001
Readmission	1		1		0.401*

Variable	Median	Mean	Median	Mean
Age(years)	37(18–75)	38.56	37(18–80)	38.38	0.863**
Duration(days)△	1(0.3-7)	1.61	1(0.2–4.5)	1.14	0.002**
Temperature(°C)	37.2(36.3–39.2)	37.23	36.8(36.2–39.1)	36.9	<0.001**
Body mass index(BMI, kg/m2)	23.06(15.83–30.20)	23.11	23.17(16.79–30.95)	23.2	0.72
White blood cell count(WBC, *109/L)	14.14(6.85–24.4)	13.98	12.78(6.78–23.54)	12.85	0.006**
Neutrophil-to-lymphocyte ratio(NLR)	8.61(2.22–30.95)	10.06	6.05(1.52–29.96)	7.3	<0.001**
C-reactive protein(CRP, mg/L)	67.4(10.8–200)	71.74	16.6(4.3–200)	26.46	<0.001**
Duration of surgery(min)	67(30–111)	66.81	52(25–101)	51.98	<0.001**
Intraoperative blood loss(ml)	10(5–50)	13.06	10(5–50)	11.66	0.054**
Length of hospital stay(LOS, days)	5(3–21)	5.27	4(2–12)	4.04	<0.001**

*Fisher’s exact test

**U test

△The medical record describes as X-(day) is set to X-0.5 (day), X+(day) is set to X + 0.5 (day)

Factors associated with RII

Overall cohort

Univariate analysis identified 10 variables associated with RII (all P < 0.05): duration ≥ 48 h, temperature ≥ 37.4 °C, CA, combined antibiotics, drainage categories, bed rest ≥ 24 h, WBC ≥ 11.76 × 10⁹/L, NLR ≥ 6.64, CRP ≥ 32.2 mg/L, and operative time ≥ 66 min. In multivariable analysis (Table 2), four factors remained independently associated with RII: CA, drainage ≥ 36 h, bed rest ≥ 24 h, and NLR ≥ 6.64. Drainage < 36 h was not significantly different from no drainage.

Table 2.

Multivariable analysis of factors associated with RII (overall cohort, n = 478)

Independent variable△	Infection(n = 59)	Noninfection (n = 419)	Univariate		Multivariate
			OR(95%CI)	P	OR(95%CI)	P
Duration(≥ 48 h)	25	103	2.256(1.286–3.958)	0.004	1.780(0.908–3.489)	0.093
Temperature(≥ 37.4 °C)	18	60	2.627(1.416–4.873)	0.002	1.421(0.654–3.085)	0.375
CA	38	70	9.002(4.993–16.300)	<0.001	4.348(1.498–12.617)	0.007
Antibiotics(Combined)	10	29	2.745(1.261–5.974)	0.018
Drainage(no/<36 h/≥36 h)	12/25/22	157/182/80	-	0.002
<36 h					1.202(0.523–2.760)	0.665
≥ 36 h					3.686(1.556–8.735)	0.003
Bed rest time(≥ 24 h)	23	73	3.028(1.694–5.413)	<0.001	3.273(1.619–6.617)	0.001
WBC(109/L)	50	242	4.063(1.947–8.480)	<0.001	2.029(0.812–5.069)	0.130
NLR	49	184	6.258(3.086–12.690)	<0.001	3.179(1.350–7.486)	0.008
CRP(mg/L)	42	123	5.945(3.258–10.848)	<0.001	1.476(0.531–4.102)	0.455
Duration of surgery(min)	26	93	2.762(1.572–4.851)	<0.001	0.984(0.469–2.064)	0.966

△Only display analysis results for variables with P < 0.05

Sensitivity analysis forcing these four variables into a reduced model yielded similar directional associations but with notably different effect estimates (CA: OR 6.18, 95%CI 3.26–11.72; drainage ≥ 36 h: OR 3.49, 95%CI 1.51–8.09; bed rest ≥ 24 h: OR 3.02, 95%CI 1.53–5.96; NLR ≥ 6.64: OR 4.95, 95%CI 2.29–10.68). This variability suggests that the effect estimates are sensitive to model specification and should be interpreted with caution, particularly given the limited number of RII events (n = 59).

Subgroup analysis by appendicitis severity

In CA patients (n = 108, RII = 38), factors independently associated with RII were bed rest ≥ 24 h, WBC ≥ 14.33 × 10⁹/L, and NLR ≥ 7.28. In UA patients (n = 370, RII = 21), only drainage ≥ 36 h was independently associated with RII (Table 3). Due to the small number of events in both subgroups (CA: 38 events, UA: 21 events), these findings should be considered exploratory.

Table 3.

Subgroup analysis of factors associated with RII (multivariable, exploratory)

Independent variable△	Univariate(CA)		Multivariate(CA)		Univariate(UA)		Multivariate(UA)
	OR(95%CI)	P	OR(95%CI)	P	OR(95%CI)	P	OR(95%CI)	P
Duration(≥ 48 h)	1.394(0.619–3.138)	0.422			2.563(1.042–6.304)	0.060*	2.256(0.867–5.869)	0.095
Drainage(no/<36 h/≥36 h)	-	0.682			-	0.001
<36 h							2.394(0.592–9.688)	0.221
≥ 36 h							7.467(1.954–28.533)	0.003
Bed rest time(≥ 24 h)	3.238(1.361–7.707)	0.007	7.212(2.221–23.416)	0.001	1.966(0.733–5.277)	0.231*
WBC(*109/L)	4.418(1.880-10.386)	<0.001	4.646(1.494–14.446)	0.008	7.266(0.962–54.899)	0.025	3.532(0.407–30.668)	0.252
NLR	8.028(2.575–25.029)	<0.001	6.024(1.617–22.448)	0.007	4.685(1.545–14.206)	0.003	2.988(0.897–9.954)	0.075
Duration of surgery(min)	1.533(0.666–3.525)	0.314			4.033(0.923–17.626)	0.046	3.431(0.759–15.503)	0.109

*Fisher’s exact test

△Only display analysis results for variables with P < 0.05

Factors associated with prolonged LOS (≥ 5 days)

Overall cohort

Univariate analysis identified seven variables associated with prolonged LOS (all P < 0.05): surgical complications, CA, drainage categories, WBC ≥ 10.19 × 10⁹/L, NLR ≥ 6.53, CRP ≥ 30.1 mg/L, and operative time ≥ 53 min. Bed rest ≥ 24 h was not associated with prolonged LOS in univariate analysis (P = 0.225). In multivariable analysis (Table 4), surgical complications and drainage ≥ 36 h were independently associated with prolonged LOS. With 198 events and 7 candidate variables, the event-per-variable ratio was sufficient, supporting the reliability of these findings.

Table 4.

Multivariable analysis of factors associated with prolonged LOS (overall cohort, n = 478)

Independent variable△	Hospitalization≥5days(n = 198)	Hospitalization<5days(n = 280)	Univariate		Multivariate
			OR(95%CI)	P	OR(95%CI)	P
Surgical complications	59	11	10.380(5.283–20.396)	<0.001	6.136(2.967–12.692)	<0.001
CA	67	41	2.981(1.914–4.644)	<0.001	1.574(0.802–3.091)	0.187
Drainage(no/<36 h/≥36 h)	50/77/71	119/130/31	-	<0.001
<36 h					1.229(0.768–1.968)	0.390
≥ 36 h					4.741(2.669–8.419)	<0.001
WBC(109/L)	168	199	2.279(1.430–3.634)	<0.001	1.691(0.977–2.926)	0.061
NLR	118	122	1.910(1.320–2.764)	0.001	1.157(0.736–1.820)	0.528
CRP(mg/L)	92	83	2.060(1.410–3.010)	<0.001	1.159(0.663–2.028)	0.605
Duration of surgery(min)	118	139	1.496(1.036–2.162)	0.032	1.214(0.791–1.864)	0.374

△Only display analysis results for variables with P < 0.05

Subgroup analysis by appendicitis severity

In CA patients (n = 108, LOS≥5d = 67), only surgical complications were independently associated with prolonged stay. In UA patients (n = 370, LOS≥5d = 131), factors independently associated with prolonged stay were surgical complications, drainage ≥ 36 h, and WBC ≥ 10.19 × 10⁹/L (Table 5). While the event numbers in these subgroups were larger than those for RII, these analyses were still considered exploratory due to the limited overall sample sizes in each subgroup.

Table 5.

Subgroup analysis of factors associated with prolonged LOS (multivariable, exploratory)

Independent variable△	Univariate(CA)		Multivariate(CA)		Univariate(UA)		Multivariate(UA)
	OR(95%CI)	P	OR(95%CI)	P	OR(95%CI)	P	OR(95%CI)	P
Surgical complications	5.004(1.947–12.862)	<0.001	4.336(1.630-11.534)	0.003	13.856(4.705–40.806)	<0.001	8.444(2.710-26.314)	<0.001
Drainage(no/<36 h/≥36 h)	-	0.402			-	<0.001
<36 h							1.017(0.590–1.752)	0.952
≥ 36 h							5.555(2.901–10.637)	<0.001
WBC(*109/L)	3.145(0.977–10.125)	0.047	1.858(0.527–6.548)	0.335	2.393(1.380–4.149)	0.002	1.969(1.028–3.772)	0.041
NLR	1.959(0.857–4.482)	0.109			1.629(1.060–2.503)	0.025	1.055(0.623–1.788)	0.841
CRP(mg/L)	0.610(0.523–0.710)	0.287*			5.117(1.333–19.632)	0.020*	3.146(0.718–13.775)	0.128

*Fisher’s exact test

△Only display analysis results for variables with P < 0.05

Discussion

In this single-center retrospective study, prolonged abdominal drainage (≥ 36 h) and bed rest (≥ 24 h) were independently associated with residual intra-abdominal infection after laparoscopic appendectomy in the overall cohort. However, the sensitivity analysis revealed that effect estimates varied notably depending on model specification, underscoring the need for cautious interpretation given the limited number of RII events. For prolonged LOS, the overall cohort analysis (with sufficient events) identified surgical complications and drainage ≥ 36 h as independent factors. Exploratory subgroup analyses revealed potential differences in risk factor profiles by appendicitis severity, which should be considered hypothesis-generating. Notably, drainage for less than 36 h was not associated with increased risk for either outcome.

Drainage duration

Recent WSES Jerusalem guidelines and meta-analyses discourage routine prophylactic drainage after appendectomy for complicated appendicitis [11, 12, 13, 21], as drainage placement itself has been associated with increased infectious complications. However, these recommendations do not address the optimal duration of drainage when placement is deemed clinically necessary. Our study extends the existing evidence by demonstrating a duration-dependent effect: drainage for ≥ 36 hours was associated with increased RII risk, whereas drainage for < 36 hours appeared safe. This nuance is clinically relevant-rather than simply asking ‘whether to drain,’ our findings address the question of ‘how long to drain’ when drainage is considered unavoidable. In this sense, our results complement current guideline recommendations by providing actionable guidance for the scenario in which the surgeon chooses to place a drain.

The exploratory subgroup analyses revealed distinct patterns by appendicitis severity. In uncomplicated appendicitis, prolonged drainage was the only modifiable factor associated with RII, with a more than sevenfold increased odds. This finding raises the possibility that in relatively clean surgical fields, drain placement may introduce pathogens or impair local defenses, outweighing any theoretical benefit. In complicated appendicitis, drainage duration was not independently associated with RII; instead, markers of infection severity (NLR, WBC) and immobility were more prominent. This could reflect that in the setting of established infection, the contribution of drainage to infectious risk is overshadowed by host and disease factors. Alternatively, the modest sample size in this subgroup may have limited the ability to detect an association.

For prolonged LOS, drainage ≥ 36 h was a significant factor in the overall cohort and in UA patients, but not in CA patients. This suggests that in complicated appendicitis, the dominant drivers of prolonged hospitalization are the complications themselves rather than drain management.

Bed rest and mobilization

Prolonged bed rest (≥ 24 h) was strongly associated with RII in the overall cohort, with an even stronger association observed in the exploratory CA subgroup analysis (OR 7.21). This observation aligns with ERAS principles and accumulating evidence that early mobilization reduces postoperative complications [14–17]. Potential mechanisms include enhanced lymphatic drainage of peritoneal contaminants, reduced postoperative ileus, improved pulmonary function, and attenuated systemic inflammation [22–25]. Experimental studies suggest that moderate physical activity may enhance immune surveillance [23]. While these findings support encouraging mobilization within 24 h after LA, the observational nature of this study precludes causal inference. Notably, bed rest was not associated with prolonged LOS, suggesting that its impact may be specific to infectious outcomes rather than overall hospital stay.

NLR and WBC as prognostic markers

Elevated NLR was independently associated with RII in the overall cohort and in the exploratory CA subgroup analysis, consistent with previous studies demonstrating its utility in predicting complicated appendicitis and postoperative complications [26, 27, 28]. Elevated WBC was associated with RII in the exploratory CA analysis and with prolonged LOS in the exploratory UA analysis, highlighting the role of systemic inflammation in both infectious and non-infectious outcomes. As readily available markers, NLR and WBC may have potential for risk stratification, though their clinical utility requires further evaluation.

Time threshold in RII definition

The 5-day time threshold for RII diagnosis was chosen pragmatically to exclude transient postoperative inflammation. Importantly, our composite RII definition included both IAA—a well-established outcome—and non-abscess infectious collections that required antimicrobial therapy. In our cohort, IAA accounted for 10 of the 59 RII events (16.9%), while the remaining 49 events represented clinically significant infections presenting as abdominal fluid collections with features suggestive of infection with associated clinical and laboratory evidence of infection. We chose this broader composite definition because non-abscess infectious collections—while less severe than IAA—nonetheless prolong recovery, require antibiotic treatment, and may delay discharge. However, we acknowledge that this definition has not been externally validated, and the inclusion of non-abscess events may introduce heterogeneity. Future studies should consider reporting IAA as a separate outcome in addition to composite infectious endpoints.

Limitations

This study has several important limitations. First, a particularly important limitation is the potential for confounding by indication. Drain placement and removal decisions were not randomized but were made at the surgeon’s discretion, influenced by intraoperative findings such as the degree of contamination, gangrenous or perforated appendix, and hemostasis concerns. Similarly, bed rest duration was influenced by the patient’s clinical condition, including pain, hemodynamic stability, and recovery trajectory. Thus, prolonged drainage and bed rest may partly reflect more severe baseline disease rather than serving as purely causal factors for RII or prolonged LOS. Although multivariable adjustment for CA, NLR, WBC, and other factors partially addressed this, residual confounding by unmeasured variables (e.g., intraoperative severity scores, surgeon experience) cannot be excluded. Second, the modest sample size for RII events (n = 59) in the overall cohort raises concern for potential overfitting. To address this, we performed a sensitivity analysis forcing only the final four variables into a reduced model. Although the direction of associations was consistent, the effect estimates differed notably between the stepwise and reduced models (e.g., CA OR ranged from 4.35 to 6.18). This variability indicates that the results are sensitive to model specification and further supports the need for cautious interpretation. These findings should be considered exploratory rather than definitive. Third, all subgroup analyses (CA and UA for both outcomes) were exploratory due to small sample sizes, particularly for RII (CA: 38 events, UA: 21 events). The wide confidence intervals observed in these subgroup analyses—notably for drainage ≥ 36 h in UA patients (OR 7.47, 95%CI 1.95–28.53)—indicate substantial imprecision and instability in the effect estimates. These findings should therefore be considered hypothesis-generating and require confirmation in larger, adequately powered studies before any subgroup-specific clinical recommendations can be made. Fourth, the composite RII definition, while clinically relevant, has not been externally validated. Fifth, dichotomization of continuous variables using ROC-derived cutoffs may inflate type I error. Sixth, the high rate of drain placement at our institution (64.6% overall, 75% in CA) may reflect local practice and limit generalizability. Finally, 4-week follow-up may miss very late infections.

Clinical implications

Despite these limitations, the findings suggest potential areas for clinical attention. Avoiding routine prophylactic drainage, particularly in uncomplicated appendicitis, may be reasonable. When drainage is deemed necessary, early removal (within 36 h) should be considered. Encouraging mobilization within 24 h postoperatively aligns with ERAS principles and may reduce infectious complications, though it did not affect LOS in this study. The exploratory findings suggesting heterogeneity in risk factors for prolonged LOS between UA and CA, if confirmed, could support tailored management: in UA, attention to drain duration and infection markers may help shorten stay; in CA, focus should be on preventing and managing complications. NLR and WBC may help identify patients who could benefit from closer monitoring. These observations require confirmation in larger, prospective studies.

Conclusion

In this exploratory study, prolonged abdominal drainage (≥ 36 h) and bed rest (≥ 24 h) were associated with an increased odds of residual intra-abdominal infection in the overall cohort. However, sensitivity analysis revealed variability in effect estimates, indicating that these findings should be interpreted with caution given the limited number of RII events. Subgroup analyses, which were exploratory, suggested potential differences in risk factors by appendicitis severity. For prolonged hospital stay, overall cohort analysis (with sufficient events) identified drainage ≥ 36 h and surgical complications as independent factors. Drainage for less than 36 h was not associated with increased risk for either outcome. These findings suggest potential benefits of avoiding routine drainage, removing drains early when used, and encouraging mobilization within 24 h. However, given the study’s limitations—particularly the potential for confounding by indication, the modest sample size for RII analysis, the variability observed in sensitivity analysis, and the exploratory nature of subgroup analyses—these results should be considered hypothesis-generating and require confirmation in larger, prospective, multicenter studies before definitive clinical recommendations can be made.

Authors’ contributions

Y.X and X.L provided the study conception. X-Z.Q、Y.X、X.L committed to collect data. Y.X and X.L performed data analysis and completed a draft manuscript. All authors reviewed the manuscript and proposed amendments, completed revision of manuscript by Y.X.

Funding

This research did not receive any funding or personal support.

Data availability

The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This retrospective cohort study was conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the Ethics Committee of Chongqing Fifth People’s Hospital (approval number: 2024CQSDWRMYYGS-002). The requirement for informed consent was waived by the Ethics Committee of Chongqing Fifth People’s Hospital because this was a retrospective study that did not involve any intervention, did not pose any risk or harm to patients, and did not involve collection of identifiable personal information. The waiver was granted in accordance with national regulations governing retrospective research using anonymized medical records.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.