Evidence on umbilical hernia repair in patients with decompensated cirrhosis, ascites, and Child–Pugh class C is limited. In this population, outcomes after emergency repair may reflect both the urgency of the presentation and the greater hepatic, renal, inflammatory, and metabolic derangement present at the time of surgery.
This retrospective single-center cohort study included all 40 eligible adults with decompensated cirrhosis, ascites, and Child–Pugh class C at admission who underwent open primary non-mesh umbilical hernia repair between January 2019 and December 2025. Twenty patients underwent elective repair, defined as definitive repair scheduled after a short period of inpatient optimization, and 20 underwent emergency repair after limited stabilization because an acute hernia-related indication made postponement unsafe. The primary outcome was all-cause 30-day mortality. Secondary outcomes were in-hospital mortality, acute kidney injury, sepsis, wound infection, persistent ascitic leakage, reoperation, length of hospital stay, and early postoperative clinical-biochemical status. Analyses were unadjusted and exploratory; no multivariable model was fitted.
At admission, the emergency group had higher total bilirubin, international normalized ratio, creatinine, leukocyte count, Child–Pugh score and MELD-Na score, and lower serum albumin and sodium (all p ≤ 0.006). During a median elective optimization interval of 4.5 (3.0–6.0) days, the MELD-Na score decreased by a median of 2 (1–3) points. On postoperative days 2–3, every reported clinical-biochemical indicator remained less favorable in the emergency group. Median hospital stay was 14.5 versus 11.0 days (p = 0.038). Thirty-day mortality was 35.0% versus 10.0% (risk difference 25.0%, 95% confidence interval −1.3 to 47.9; p = 0.127).
Patients undergoing emergency repair reached surgery with substantially greater clinical and biochemical derangement, had a less favorable early postoperative profile and stayed longer in the hospital, while the mortality difference did not reach statistical significance. Because operative timing was determined by the clinical course, the groups were not comparable at baseline and no causal effect of urgency or of preoperative optimization can be inferred. The findings support early multidisciplinary assessment while an elective operative pathway remains feasible.
What is not yet known on the issue addressed in the submitted manuscript
In patients with decompensated cirrhosis, ascites and Child–Pugh class C, it remains unclear to what extent the less favorable outcomes after emergency umbilical hernia surgery reflect urgency itself or the greater biological severity with which these patients present for surgery.
The research hypothesis
We hypothesize that patients with decompensated cirrhosis, ascites, and Child–Pugh class C who require emergency umbilical hernia repair present for surgery in a more severe biological condition and have a less favorable early postoperative course than those operated on electively after short preoperative optimization.
The novelty added by manuscript to the already published scientific literature
The study compares two equally-sized groups sharing the same Child–Pugh class at admission, the same hernia location, and the same primary non-mesh repair technique, while documenting a short preoperative optimization window during which the MELD-Na score decreased in this very high-risk subgroup.
Ascites is a defining manifestation of decompensated liver cirrhosis and contributes to the development of umbilical hernia through chronically increased intra-abdominal pressure, progressive weakening of the abdominal wall, and impaired tissue healing [1–4]. In patients with cirrhosis, an umbilical hernia is therefore more than an isolated abdominal wall defect. Its course may be complicated by incarceration, strangulation, overlying skin ulceration or necrosis, spontaneous rupture, ascitic fluid leakage, peritonitis, and evisceration [2–5].
Historically, the substantial perioperative risk associated with advanced cirrhosis favored conservative management. More recent evidence indicates that emergency repair is associated with worse outcomes than elective repair in appropriately selected patients [3–10]. This is particularly relevant in patients with persistent ascites, advanced hepatic dysfunction, renal impairment, hyponatremia, infection and coagulation abnormalities, each of which may adversely affect the perioperative outcome [1, 11].
The interpretation of this association remains difficult in patients with the most advanced liver disease. Published cohorts commonly span Child–Pugh classes A, B and C, which limits the direct application of their conclusions to patients in class C [6–10]. Moreover, patients requiring emergency repair usually present with an acute local process and more profound systemic decompensation, whereas an elective procedure is possible only when the clinical situation permits at least a short period of preoperative preparation. Operative timing is therefore closely linked to baseline severity, and any comparison between the two settings is subject to confounding by indication.
We hypothesized that, within a cohort restricted to decompensated cirrhosis, ascites, and Child–Pugh class C at admission, patients undergoing emergency umbilical hernia repair would have greater biological derangement at the time of surgery and a less favorable early postoperative course than patients undergoing elective repair after short preoperative optimization. The aim of the study was to compare baseline clinical and biochemical characteristics, perioperative dynamics, and early outcomes between these two clinically defined groups. The comparison was intended to be descriptive and was not designed to estimate an independent causal effect of emergency surgery.
This was a retrospective, single-center cohort study conducted at the Department of Surgery No. 2 of Nicolae Testemițanu State University of Medicine and Pharmacy, based at the Holy Trinity Clinical Municipal Hospital, Chișinău, Republic of Moldova. The source population comprised adult patients who underwent surgery for an umbilical hernia between January 2019 and December 2025. All eligible records identified during the study period were included; there was no prospective enrollment and no sampling of eligible cases. Data were abstracted from admission records, inpatient charts, operative reports, laboratory results, and postoperative monitoring documentation and entered into a study database. The study is reported in accordance with the STROBE statement for cohort studies [12].
Inclusion criteria were age 18 years or older; an acquired umbilical hernia; decompensated liver cirrhosis; clinically evident ascites; Child–Pugh class C at admission; and treatment by open primary umbilical hernia repair without mesh. Exclusion criteria were non-cirrhotic ascites, Child–Pugh class A or B at admission, inguinal or incisional hernia, recurrent hernia, another primary ventral hernia, concomitant malignancy, or major intra-abdominal septic disease requiring a different principal operation. Forty patients met these criteria.
Patients were classified according to the timing of definitive repair, which in turn was determined by whether an acute hernia-related indication required immediate operation. The elective group comprised 20 patients in whom definitive repair could be scheduled after a short period of inpatient optimization, because no indication for immediate surgery persisted after initial assessment and conservative local management. The emergency group comprised 20 patients who underwent repair after only limited stabilization because an acute hernia-related indication made postponement unsafe.
Acute hernia-related indications for immediate operation were incarceration that remained irreducible after analgesia and, where appropriate, decompressive paracentesis; strangulation; intestinal obstruction; spontaneous rupture of the hernia with uncontrolled ascitic leakage; and full-thickness cutaneous necrosis with impending rupture over a tense hernia [13, 14]. Cutaneous changes overlying the hernia were recorded as a separate variable, because superficial erosion or ulceration without necrosis, and ascitic leakage that ceased with conservative measures, were not in themselves indications for immediate surgery. Conversely, incarceration, strangulation, and intestinal obstruction could mandate urgent operation in the absence of any cutaneous abnormality. The local findings at presentation and the principal indication that determined the timing of repair in each group are reported in Table 2, which resolves the apparent discordance between cutaneous status and operative setting.
At admission, clinical examination assessed hernia reducibility, pain, local skin changes, ascitic fluid leakage, intestinal obstruction, and peritoneal signs. Laboratory assessment included full blood count, coagulation profile, total bilirubin, serum albumin, creatinine, sodium, and routine biochemistry. Abdominal ultrasonography was used to assess ascites and intra-abdominal findings. Ascitic fluid analysis included macroscopic appearance, albumin, total protein, glucose, lactate dehydrogenase, total leukocyte, and polymorphonuclear neutrophil (PMN) counts and bacteriological culture. The reported ascitic-fluid variables were available for all 40 patients.
The severity of liver disease was assessed with the Child–Turcotte–Pugh score [15] and the Model for End-Stage Liver Disease with sodium (MELD-Na) score [16, 17]. An ascitic PMN count of at least 250/mm³, in the absence of a surgically treatable intra-abdominal source, was interpreted as neutrocytic ascites consistent with spontaneous bacterial peritonitis [1, 11, 18]. Positive ascitic cultures were recorded separately, because monomicrobial bacteriascites may occur with a PMN count below 250/mm³ [19].
The primary outcome was all-cause mortality within 30 days of surgery, the standard short-term surgical mortality endpoint. In-hospital mortality was retained as a secondary outcome because it describes deaths during the index admission and may diverge from 30-day mortality when patients are discharged earlier or remain hospitalized beyond 30 days; in the present cohort, one death occurred after discharge but within 30 days. Other secondary outcomes were acute kidney injury, sepsis, wound infection, persistent ascitic leakage, reoperation, length of hospital stay, and the early postoperative clinical-biochemical profile. Non-fatal postoperative outcomes and reoperations were ascertained during the index hospital admission.
For patients who remained in the hospital through postoperative day 30, vital status was established from inpatient records. For those discharged earlier, vital status was ascertained from existing clinical documentation generated during routine postoperative care, including documented telephone contact by the treating team, outpatient review, and hospital readmission records. No patient was contacted specifically for this retrospective analysis. Thirty-day vital status was available for all 40 patients. Acute kidney injury was defined according to the International Club of Ascites criteria for cirrhosis (ICA-AKI) [20], and sepsis according to the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [21]. Persistent ascitic leakage was defined as loss of ascitic fluid through the operative wound lasting more than 48 hours after surgery.
Patients in the elective group received an individualized period of optimization directed at ascites, hepatic decompensation, and major fluid, electrolyte, infectious and coagulation abnormalities. Because management was reconstructed retrospectively and was not delivered through a single standardized protocol, the dose and intensity of every component could not be quantified uniformly. Depending on clinical need and tolerance, management included sodium and volume correction, diuretic therapy, albumin administration, treatment of suspected or documented infection, lactulose for hepatic encephalopathy, and correction of major coagulation or electrolyte abnormalities.
Therapeutic paracentesis was not routine. It was used selectively in patients with tense ascites, marked abdominal wall tension, respiratory discomfort, or symptomatic abdominal distension, with albumin replacement according to the treating team’s decision. The decision to proceed with elective repair was taken after confirmation that no acute local indication persisted and after partial clinical and biological stabilization, with attention to bilirubin, international normalized ratio (INR), serum sodium, leukocyte count, creatinine, ascites control, and encephalopathy. Patients in the emergency group received only limited stabilization, focused on hemodynamic support, analgesia, rapid correction of immediately relevant abnormalities, and antibacterial treatment when infection was suspected or documented.
All patients underwent open primary tissue repair of the umbilical hernia with monofilament suture and without prosthetic mesh. This was the operative technique used in both groups. In emergency cases, additional maneuvers were performed when required by the operative findings, including reduction of hernial contents, assessment of intestinal viability, and bowel resection when a non-viable segment was identified.
Postoperative monitoring focused on hemodynamic status, renal function, sodium balance, wound condition, ascites, infection, and further hepatic decompensation. Abdominal drainage was used selectively according to intraoperative findings and the estimated risk of fluid accumulation or ascitic leakage. Postoperative ascites management was individualized and included sodium restriction, diuretic therapy when tolerated, albumin when indicated, therapeutic paracentesis for recurrent tense ascites, renal protection, and wound surveillance.
For the analysis of biological dynamics, the elective group was assessed at admission, immediately before surgery after optimization, and on postoperative days 2–3. The emergency group was assessed at admission, which also served as the immediately preoperative assessment because only limited stabilization was possible, and on postoperative days 2–3.
Analyses were performed with IBM SPSS Statistics, version 29.0 (IBM Corp., Armonk, NY, USA). Continuous variables are reported as median and interquartile range (IQR), and categorical variables as number and percentage. Between-group comparisons used the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables. Paired admission and immediately preoperative measurements in the elective group were compared with the Wilcoxon signed-rank test. For binary outcomes, the absolute risk difference between groups is additionally reported with a 95% confidence interval (CI) calculated by the Newcombe hybrid score method [22], in order to convey the precision of the estimate. All tests were two-sided, and p < 0.05 was considered statistically significant.
The analyses were exploratory. No correction for multiple comparisons was applied, and individual p-values should therefore be interpreted cautiously. Only nine deaths occurred within 30 days, and the two groups were markedly imbalanced at baseline. A multivariable logistic regression model would consequently have been unstable and potentially misleading, and no adjusted model was fitted. Results are presented as unadjusted descriptive comparisons and do not estimate an independent treatment effect.
The study was approved by the Research Ethics Committee of Nicolae Testemițanu State University of Medicine and Pharmacy, Chișinău, Republic of Moldova (approval no. 1 of 12.05.2022, following the review of application no. 8 of 10.01.2022). The study was conducted in accordance with the Declaration of Helsinki. Patients were not prospectively enrolled, and no study-specific informed consent was sought after treatment; the retrospective analysis of anonymized records was performed under the ethics approval cited above. The dataset was assembled exclusively from existing clinical records, including routine postoperative follow-up entries; no patient was contacted for research purposes, no identifying information is reported, and only anonymized data were analyzed.
The final cohort comprised 40 patients: 20 in the elective group and 20 in the emergency group. Every patient had decompensated cirrhosis, clinically evident ascites, and Child–Pugh class C at admission. Age and sex distribution did not differ significantly between groups.
Baseline demographic, clinical and biochemical characteristics are shown in Table 1. Hernia-related cutaneous changes at initial presentation were documented in 4 patients (20.0%) in the elective group and 15 (75.0%) in the emergency group (risk difference 55.0%, 95% CI 24.3 to 73.3; p = 0.001). At admission, the emergency group also had significantly higher total bilirubin, INR, serum creatinine, leukocyte count, Child–Pugh score, and MELD-Na score, together with lower serum albumin and sodium. Hepatic encephalopathy, an ascitic PMN count of at least 250/mm³, and a positive ascitic culture were proportionally more frequent in the emergency group, but the corresponding differences were not statistically significant. These findings demonstrate substantial baseline non-equivalence between the groups.
Table 1. Baseline characteristics at admission | |||
Variable | Elective (n=20) | Emergency (n=20) | p |
Age, years | 57 (53–62) | 60 (58–66) | 0.184 |
Male sex, n (%) | 14 (70.0) | 16 (80.0) | 0.716 |
Hernia-related cutaneous changes at presentation, n (%) | 4 (20.0) | ||
The local findings at presentation and the principal indication that determined the timing of repair are shown in Table 2. Cutaneous changes were present in four elective patients, in whom the local findings were superficial and no acute indication for immediate operation persisted after initial assessment, so that a short inpatient preparation remained possible. Conversely, five emergency patients had no cutaneous abnormality but required urgent operation for other acute hernia-related indications.
Table 2. Local hernia findings and principal indication for the timing of repair | ||
Finding / indication at presentation | Elective (n=20) | Emergency (n=20) |
A. Local findings (categories not mutually exclusive), n (%) | ||
Reducible hernia, no cutaneous change | 16 (80.0) | 0 (0.0) |
Cutaneous erosion or ulceration without necrosis | 4 (20.0) | 9 (45.0) |
Full-thickness cutaneous necrosis | 0 (0.0) | 6 (30.0) |
The median duration of preoperative optimization in the elective group was 4.5 (IQR 3.0–6.0) days. Paired comparisons between admission and the immediately preoperative assessment showed statistically significant changes in total bilirubin, serum albumin, INR, serum sodium, leukocyte count, Child–Pugh score, and MELD-Na score. Serum creatinine decreased numerically, but the paired difference was not statistically significant (p = 0.117). The median decrease in MELD-Na was 2 (IQR 1–3) points. These within-group changes were observed during the optimization interval and cannot be attributed causally to any individual component of management. Immediately before surgery, every reported laboratory or prognostic parameter remained significantly less favorable in the emergency group (Table 3).
Table 3. Clinical-biochemical changes in the elective group and preoperative comparison between groups. | |||||
Variable | Elective, admission (n=20) | Elective, preoperative (n=20) | p¹ | Emergency, admission/preop. (n=20) | p² |
Total bilirubin, mg/dL | 4.75 (4.08–5.39) | 4.33 (3.75–5.22) | 0.044 | 6.27 (5.98–6.66) | <0.001 |
Serum albumin, g/dL | 2.42 (2.34–2.59) | ||||
Ascitic fluid findings at admission are shown in Table 4. Compared with the elective group, the emergency group had lower ascitic albumin and glucose concentrations and higher lactate dehydrogenase, total leukocyte, and PMN counts. Total ascitic protein and the serum–ascites albumin gradient did not differ significantly, and the proportion of patients with an ascitic PMN count of at least 250/mm³ was 25.0% versus 35.0% (p = 0.731).
Table 4. Ascitic fluid profile at admission. | |||
Variable | Elective (n=20) | Emergency (n=20) | p |
Ascitic albumin, g/dL | 0.98 (0.78–1.15) | 0.57 (0.45–0.70) | <0.001 |
Total ascitic protein, g/dL | 1.52 (1.23–1.95) | 1.52 (1.27–1.69) | 0.709 |
Ascitic glucose, mg/dL | 93.0 (83.8–103.3) | ||
On postoperative days 2–3, the emergency group continued to have significantly higher total bilirubin, INR, creatinine, leukocyte count, Child–Pugh score, and MELD-Na score, and lower serum albumin and sodium, than the elective group (Table 5). The baseline severity difference therefore remained evident during the early postoperative period.
Table 5. Early postoperative clinical-biochemical status on postoperative days 2–3. | |||
Variable | Elective (n=20) | Emergency (n=20) | p |
Total bilirubin, mg/dL | 4.85 (4.31–5.48) | 7.70 (7.22–8.35) | <0.001 |
Serum albumin, g/dL | 2.32 (2.23–2.51) | 1.89 (1.60–1.99) | <0.001 |
INR | 2.05 (1.83–2.20) | ||
Postoperative outcome rates are presented in Table 6. Acute kidney injury occurred in 20.0% of elective and 45.0% of emergency patients, and sepsis in 10.0% and 35.0%, respectively. Wound infection, persistent ascitic leakage, and reoperation were also proportionally more frequent after emergency repair, but none of these comparisons reached statistical significance and all corresponding confidence intervals for the risk difference included zero. Median hospital stay was significantly longer in the emergency group: 14.5 (IQR 11.5–16.0) versus 11.0 (9.0–13.3) days (p = 0.038). In-hospital mortality was 30.0% after emergency and 10.0% after elective repair (p = 0.235). Thirty-day mortality, the primary outcome, was 35.0% and 10.0%, respectively (risk difference 25.0%, 95% CI −1.3 to 47.9; p = 0.127). No adjusted effect estimate was calculated.
Table 6. Postoperative outcomes. | ||||
Outcome | Elective (n=20) | Emergency (n=20) | Risk difference, % (95% CI) | p |
Acute kidney injury, n (%) | 4 (20.0) | 9 (45.0) | 25.0 (−3.9 to 49.0) | 0.176 |
Sepsis, n (%) | 2 (10.0) | 7 (35.0) | 25.0 (−1.3 to 47.9) | 0.127 |
The principal finding of this study is that the elective and emergency groups were clinically different at baseline despite every patient being in Child–Pugh class C at admission. Patients who required emergency repair reached surgery with greater hepatic, renal, electrolyte, and inflammatory derangement and remained more severely decompensated on postoperative days 2–3, while patients undergoing elective repair showed several favorable within-group changes during a short preoperative interval. These results support the study hypothesis at a descriptive level, but they do not demonstrate that operative timing itself caused the observed differences.
This baseline non-equivalence is the central interpretative constraint of the study and represents an unavoidable limitation of this retrospective comparison. Emergency repair was performed precisely in those patients whose local condition left no time for optimization, and elective repair was possible precisely because no acute indication persisted. The adverse baseline profile of the emergency group therefore reflects the combined burden of acute local disease, uncontrolled ascites, inflammation or infection, impaired effective circulation, and renal dysfunction [1, 11]. Confounding by indication of this magnitude cannot be removed by between-group testing, nor, in a cohort with nine deaths, by statistical adjustment. For the same reason, we deliberately refrained from fitting a multivariable model and confined inference to observed event rates, unadjusted comparisons, and interval estimates.
The separation of cutaneous findings from the indication for immediate surgery, reported in Table 2, is relevant to this interpretation. Skin changes overlying a tense hernia were recorded in both groups and were not, by themselves, decisive: superficial erosion or ulceration and leakage that ceased under conservative management still permitted a period of preparation, whereas incarceration, strangulation, or obstruction mandated immediate operation even when the overlying skin was intact. Reporting these categories separately avoids the assumption that cutaneous status and operative urgency are interchangeable classifications and explains why four elective patients had local skin changes, and five emergency patients did not.
The unadjusted 30-day mortality was 35.0% after emergency repair and 10.0% after elective repair. The direction of this difference is consistent with published evidence, although the comparison in the present cohort did not reach statistical significance and the confidence interval for the risk difference was wide. In a national Veterans Affairs cohort, Johnson et al. reported 30-day mortality of 12.2% after emergency repair and 1.2% after non-emergency repair, with a further increase in risk among patients with ascites [8]. In an English population-based study, Adiamah et al. reported 90-day case fatality of 19% after emergency repair and 2% after elective repair in patients with cirrhosis, with greater risk in decompensated disease [10]. A systematic review including 3,229 patients found higher mortality after emergency than elective repair, with a pooled odds ratio of 2.67 (95% CI 1.87–3.97) [9].
The higher absolute mortality observed in both of our groups is clinically plausible, because the cohort was intentionally restricted to patients with ascites and Child–Pugh class C, whereas population-based studies include a broader spectrum of liver disease. The present series should therefore be read as an estimate of the event rates that may be expected in the most severely affected subgroup, not as an independent test of the effect of urgency.
During the median 4.5-day preoperative interval in the elective group, bilirubin, albumin, INR, sodium, leukocyte count, Child–Pugh score, and MELD-Na score changed significantly, and the median MELD-Na score decreased by 2 points. These observations show that short-term biological change was measurable in selected patients, but they should not be read as evidence that the optimization package caused the changes or improved survival. The intervention was individualized, its components were not quantified uniformly, and no comparison group with a similar initial condition was managed without preparation. The data therefore describe the clinical course during a preparation window rather than the efficacy of a standardized optimization protocol.
The early postoperative laboratory data demonstrate persistence of the initial severity gap. On days 2–3 the emergency group remained worse across every reported hepatic, renal, inflammatory, and prognostic measure, so that surgery did not eliminate the biological differences present beforehand. This is relevant when counseling patients and planning postoperative surveillance, because the operation takes place within an already fragile systemic state rather than in isolation. The individual adverse-outcome rates followed the same direction: acute kidney injury, sepsis, wound infection, persistent ascitic leakage, and reoperation were each proportionally more frequent after emergency repair, although no individual comparison was statistically significant and every corresponding confidence interval crossed zero. Advanced cirrhosis is characterized by a fragile hemodynamic balance, and urgent surgery in the presence of infection, fluid shifts, and uncontrolled ascites may be accompanied by renal deterioration and impaired wound healing [1, 11]. Among the clinical outcomes in Table 6, only length of hospital stay differed significantly; given the sample size and the number of comparisons, the remaining differences should be regarded as imprecise descriptive signals rather than evidence of a treatment effect.
A strength of the study is its focus on a narrowly defined and particularly high-risk population that is underrepresented in the literature. All patients had the same hernia location, ascites, and Child–Pugh class C at admission and underwent the same primary non-mesh operative technique. Dynamic measurements at admission, immediately before elective repair, and on postoperative days 2–3 add information that cannot be obtained from a single perioperative assessment.
Several limitations require emphasis. First, the retrospective single-center design limits generalizability and permits residual information bias. Second, a sample of 40 patients with nine 30-day deaths provides low statistical power for mortality and other infrequent outcomes, and the resulting estimates are imprecise. Third, the groups differed substantially at admission, reflecting strong confounding by indication, so that the data cannot determine the independent effect of emergency repair or of the preparation interval. Fourth, the individual components and intensity of preoperative management could not be reconstructed uniformly from the records. Fifth, the analysis comprised multiple exploratory comparisons without multiplicity correction. Sixth, follow-up was restricted to 30 days, so hernia recurrence, longer-term survival, and quality of life could not be assessed, and non-fatal events after discharge were not captured systematically. Seventh, the outcome definitions were applied to information recorded during routine clinical care rather than to data collected prospectively for research; wound infection, sepsis, and acute kidney injury were therefore ascertained from the documentation available in the medical records. Finally, because the study describes a single institutional practice, the observed optimization interval may not be reproducible in settings with different admission thresholds or resources.
These limitations notwithstanding, the results carry a practical message. An umbilical hernia in a patient with decompensated cirrhosis and ascites should prompt early surgical, hepatological, and anesthetic assessment rather than indefinite deferral. The objective is not to operate on every patient immediately, but to identify the interval in which local findings still permit preparation and planned repair. Prospective multicenter studies using standardized definitions, documented optimization pathways, and sufficient event numbers are needed to determine whether such an approach improves outcomes.
Among patients with decompensated cirrhosis, ascites, and Child–Pugh class C, those undergoing emergency umbilical hernia repair reached surgery with substantially greater clinical and biochemical derangement than those undergoing elective repair, remained more decompensated on postoperative days 2–3, and stayed significantly longer in hospital. Thirty-day mortality was 35.0% versus 10.0%, a difference that did not reach statistical significance.
In selected patients admitted in Child–Pugh class C, a short preoperative interval with measurable biological change was nevertheless attainable before repair, indicating that the decisive clinical question is not only whether an operation is elective or urgent, but whether such an interval still exists when the patient is first assessed. These observations support early multidisciplinary evaluation while an elective operative pathway remains feasible.
Not declared.
SP and GA conceived the study and designed the analytical framework. SP, GA, OC, TZ, GL, and LC collected the clinical data. SP and GA performed data analysis and interpretation. SP drafted the initial version of the manuscript. All authors critically reviewed the manuscript for important intellectual content, read and approved the final version, and agree to be accountable for all aspects of the work.
The study was conducted within the postdoctoral scientific project "Pathogenetic aspects and particularities in the surgical treatment of ascitic syndrome in patients with decompensated liver cirrhosis", project code 23.00208.8007.02/PD I, Strategic Priority: Health (National Agency for Research and Development (NARD / ANCD) of the Republic of Moldova). The funder had no role in the study design, data collection, analysis and interpretation, or manuscript writing.
The study was approved by the Research Ethics Committee of Nicolae Testemițanu State University of Medicine and Pharmacy, Chișinău, Republic of Moldova (approval no. 1 of 12.05.2022, following the review of application no. 8 of 10.01.2022). The research was conducted in accordance with the principles of the Declaration of Helsinki.
Obtained.
Not commissioned, externally peer reviewed.
The authors used ChatGPT to assist with limited language refinement. The authors critically reviewed, revised, and approved the final version of the manuscript and take full responsibility for its content.
Sergiu Pisarenco – https://orcid.org/0009-0003-6516-1630
Gheorghe Anghelici – https://orcid.org/0009-0003-1063-2802
Tatiana Zugrav – https://orcid.org/0000-0002-5205-5471
Oleg Crudu – https://orcid.org/0009-0003-8616-7222
Gheorghe Lupu – https://orcid.org/0009-0004-0901-0341
Liviu Chiriac – https://orcid.org/0009-0005-0854-1295
15 (75.0)
0.001 |
Hepatic encephalopathy at admission, n (%) | 8 (40.0) | 11 (55.0) | 0.527 |
Ascitic PMN ≥250/mm³ at admission, n (%) | 5 (25.0) | 7 (35.0) | 0.731 |
Positive ascitic culture, n (%) | 1 (5.0) | 4 (20.0) | 0.342 |
Total bilirubin, mg/dL | 4.75 (4.08–5.39) | 6.27 (5.98–6.66) | <0.001 |
Serum albumin, g/dL | 2.42 (2.34–2.59) | 2.13 (2.03–2.25) | <0.001 |
INR | 1.99 (1.83–2.07) | 2.34 (2.15–2.42) | <0.001 |
Serum creatinine, mg/dL | 1.26 (1.15–1.50) | 1.64 (1.39–1.82) | 0.001 |
Serum sodium, mmol/L | 131.0 (129.0–131.3) | 127.5 (126.0–129.3) | <0.001 |
Leukocytes, ×10⁹/L | 7.70 (6.88–9.82) | 10.95 (10.10–12.90) | <0.001 |
Child–Pugh score | 12 (12–13) | 13 (13–14) | 0.006 |
MELD-Na score | 26 (24–28) | 31 (30–32) | <0.001 |
Note: Data are presented as median (interquartile range) for continuous variables and number (%) for categorical variables. The row “hernia-related cutaneous changes” records local skin findings independently of the indication for immediate surgery and must not be read as the total number of complicated hernias; the corresponding breakdown is given in Table 2. INR – international normalized ratio; PMN – polymorphonuclear neutrophils; MELD-Na – Model for End-Stage Liver Disease with sodium. Used statistical analysis: Mann–Whitney U test (continuous variables) and Fisher’s exact test (categorical variables); p < 0.05 was considered statistically significant. |
Ascitic leakage through the hernia
2 (10.0) |
8 (40.0) |
Irreducibility at first examination | 1 (5.0) | 11 (55.0) |
B. Principal indication determining timing (mutually exclusive), n (%) |
No acute indication – repair scheduled after optimization | 16 (80.0) | 0 (0.0) |
Cutaneous change resolved or controlled conservatively – repair scheduled | 4 (20.0) | 0 (0.0) |
Irreducible incarceration | 0 (0.0) | 6 (30.0) |
Strangulation | 0 (0.0) | 3 (15.0) |
Intestinal obstruction | 0 (0.0) | 2 (10.0) |
Spontaneous rupture with uncontrolled ascitic leakage | 0 (0.0) | 5 (25.0) |
Cutaneous necrosis with impending rupture | 0 (0.0) | 4 (20.0) |
C. Intraoperative bowel resection performed, n (%) |
Bowel resection performed | 0 (0.0) | 4 (20.0) |
Note: Panel A describes local findings at first examination and the categories are not mutually exclusive; a patient could present with more than one finding. Panel B reports the single principal indication that determined whether repair was performed after a period of optimization or immediately, and the categories sum to 20 in each group. Used statistical analysis: descriptive; no formal significance testing was applied to these subcategories because of the small numbers per cell. |
2.60 (2.48–2.71)
<0.001 |
2.13 (2.03–2.25) |
<0.001 |
INR | 1.99 (1.83–2.07) | 1.83 (1.69–1.99) | <0.001 | 2.34 (2.15–2.42) | <0.001 |
Serum creatinine, mg/dL | 1.26 (1.15–1.50) | 1.17 (1.12–1.46) | 0.117 | 1.64 (1.39–1.82) | <0.001 |
Serum sodium, mmol/L | 131.0 (129.0–131.3) | 132.0 (130.0–134.0) | <0.001 | 127.5 (126.0–129.3) | <0.001 |
Leukocytes, ×10⁹/L | 7.70 (6.88–9.82) | 5.95 (4.47–8.05) | <0.001 | 10.95 (10.10–12.90) | <0.001 |
Child–Pugh score | 12 (12–13) | 12 (11–13) | 0.024 | 13 (13–14) | 0.001 |
MELD-Na score | 26 (24–28) | 24 (22–26) | <0.001 | 31 (30–32) | <0.001 |
Note: Data are presented as median (interquartile range). The emergency group had no optimization interval, so its admission measurement also served as the immediately preoperative measurement. Abbreviations as in Table 1. Statistical methods: p¹ – paired comparison between admission and preoperative values in the elective group (Wilcoxon signed-rank test); p² – comparison between the preoperative values of the elective group and the admission/preoperative values of the emergency group (Mann–Whitney U test); p < 0.05 was considered statistically significant. |
69.5 (57.3–78.3)
<0.001 |
Ascitic LDH, U/L | 113.5 (87.5–147.0) | 258.0 (198.0–353.3) | <0.001 |
Total ascitic leukocytes, /mm³ | 219.0 (147.8–279.0) | 635.0 (293.0–1671.8) | 0.002 |
Ascitic PMN, /mm³ | 97.5 (50.0–186.5) | 234.5 (124.0–777.5) | 0.013 |
Ascitic PMN ≥250/mm³, n (%) | 5 (25.0) | 7 (35.0) | 0.731 |
SAAG, g/dL | 1.35 (1.16–1.55) | 1.56 (1.30–1.82) | 0.089 |
Note: Data are presented as median (interquartile range) or number (%). LDH – lactate dehydrogenase; PMN – polymorphonuclear neutrophils; SAAG – serum–ascites albumin gradient. Statistical methods: Mann–Whitney U test (continuous variables) and Fisher’s exact test (categorical variable); p < 0.05 was considered statistically significant. |
2.67 (2.47–2.87)
<0.001 |
Serum creatinine, mg/dL | 1.29 (1.21–1.52) | 1.94 (1.65–2.60) | <0.001 |
Serum sodium, mmol/L | 130.5 (128.0–132.3) | 125.0 (122.0–128.0) | <0.001 |
Leukocytes, ×10⁹/L | 9.45 (7.15–12.08) | 16.80 (14.23–19.43) | <0.001 |
Child–Pugh score | 12 (12–13) | 14 (13–14) | 0.001 |
MELD-Na score | 26 (24–29) | 34.5 (33.0–37.3) | <0.001 |
Note: Data are presented as median (interquartile range), measured on postoperative days 2–3. Abbreviations as in Table 1. Statistical methods: Mann–Whitney U test; p < 0.05 was considered statistically significant. |
Wound infection, n (%) | 2 (10.0) | 4 (20.0) | 10.0 (−13.4 to 32.8) | 0.661 |
Persistent ascitic leakage, n (%) | 4 (20.0) | 8 (40.0) | 20.0 (−8.2 to 44.5) | 0.301 |
Reoperation, n (%) | 1 (5.0) | 4 (20.0) | 15.0 (−7.1 to 37.0) | 0.342 |
Length of hospital stay, days | 11.0 (9.0–13.3) | 14.5 (11.5–16.0) | – | 0.038 |
In-hospital mortality, n (%) | 2 (10.0) | 6 (30.0) | 20.0 (−5.4 to 43.1) | 0.235 |
30-day mortality, n (%) | 2 (10.0) | 7 (35.0) | 25.0 (−1.3 to 47.9) | 0.127 |
Note: Data are presented as number (%) or median (interquartile range) for length of hospital stay. Risk difference is the absolute difference in event rate (emergency minus elective) with a 95% confidence interval (CI) calculated using the Newcombe hybrid score method; it is not calculated for continuous variables. Acute kidney injury was defined by ICA-AKI criteria and sepsis by Sepsis-3 criteria. Statistical methods: Fisher’s exact test (categorical outcomes) and Mann–Whitney U test (length of hospital stay); p < 0.05 was considered statistically significant. |