Nationwide analysis of prolonged hospital stay and readmission after elective ventral hernia repair

No detailed large-scale data are available to explain prolonged length of hospital stay (LOS) and readmission following ventral hernia repair. Recent nationwide Danish data from the 2005-2006 period have shown an unsatisfactory 30-day outcome following elective incisional hernia repair [1]. The study comprised 2,896 incisional hernia repairs and showed that the postoperative morbidity, readmission and mortality rates were 11%, 11% and 0.4%, respectively. Ten percent of the patients stayed in hospital for > 6 days after surgery [1]. National Danish data following elective and predominantly small umbilical and epigastric hernia repairs (n = 3,431) have shown low morbidity and mortality rates (4% and 0.1%, respectively), but an unexpectedly high readmission rate (5%) [2]. No differences in outcome after incisional and umbilical/epigastric hernia repair were found between open or laparoscopic procedures [1].

The aim of the present study was to analyse factors associated with prolonged LOS, readmission and death in a nationwide group of patients undergoing elective ventral hernia repair in search for areas of interest to improve outcome after ventral hernia surgery.

MATERIAL AND METHODS

This was a descriptive Danish case-control study based on prospective results after elective ventral hernia repair during 2008. Data comprised all elective ventral hernias registered in the Danish Ventral Hernia Database (DVHD) and Danish National Patient Register (DNPR) (study cohort, Figure 1). Ventral hernias were defined as incisional, umbilical/epigastric, parastomal or other rare hernia repairs. The perioperative data in the DVHD are registered online by the operating surgeon immediately after the operation [3].The prospective data from the DVHD were matched with data from the DNPR using each patient’s social security number. By combining data from the two databases, we obtained information from the DNPR (gender, age, LOS, readmission, other procedures performed and contacts to the health-care system), as well as technical information about the ventral hernia repair from the DVHD (type of hernia, laparoscopic or open repair, use of mesh, size (wide and length) of hernia and mesh, type of mesh or suture, mesh fixation, emergency or elective repair, primary or recurrent repair). Data (study group) were supplemented with retrospective information from manual analyses of patients’ medical records. The exclusion criteria were emergency repairs and hernia repairs with additional surgical procedures unrelated to the hernia repair.

The study group included patients with poor outcome defined as LOS ≥ 5 days and/or readmission and death within 30 days. The control group were patients without poor outcome (Figure 1).

LOS was defined as the period from surgery to discharge from hospital (1 day = 24 hours). LOS during readmission was not included in the calculation of postoperative LOS. Ventral hernia repairs were divided into incisional, umbilical/epigastric, parastomal and a group of rare hernia repairs (Spigelian, lumbar, trocar, etc.). The size of the hernial defect was defined as the widest described diameter (cm) of the defect.

Patients’ medical records (study group) were analysed according to well-defined predetermined parameters: The existence of co-morbidity in the study group was defined as a medically treated disease (we had no information of co-morbidity in the control group). Major perioperative complications were defined as severe and potentially fatal complications (intraoperatively recognized visceral perforation, thromboembolic complications, intensive care treatment) or as complications requiring emergency reoperation, excluding superficial skin opening, skin puncture and skin drainage. Minor perioperative complications were other complications not included in the above definitions. Pain causing prolonged LOS was defined as pain treated with intravenous opioids. If causes for prolonged LOS were not obvious in the hospital files, consensus was obtained through discussion between the investigators to identify reasons for prolonged LOS. In cases in which the LOS or readmission was not explained by a major or minor complication, we registered the other explanations entered into the medical records.

Statistics

The morbidity and mortality rates were based on the number of hernia repairs and number of patients, respectively. To prevent statistical distortion of morbidity results, we assessed only one complication (the most severe) per hernia repair [4]. Data are presented as exact numbers and percentages. Confidence intervals for proportions (95%) were calculated between the study group and control group for the specific hernia repairs. Non-parametric statistics was used in order to analyse differences in hernia sizes (Mann-Whitney) and chi square test was used to compare categorical data between the study group and the control group. A p value < 0.05 was considered statistically significant.

Trial registration: The study was approved by the Danish Data Protection Agency (ref. no. 2008-58-0020) and registered in www.clinicaltrials.gov (ref. no. NCT01388634).

RESULTS

The study profile is shown in Figure 1 and patients’ characteristics are given in Table 1 and Table 2. Overall, the study cohort contained 2,258 patients (2,275 repairs). The study group and the control group consisted of 258 patients (259 repairs) and 2,000 patients (2,016 repairs), respectively. In total, 132 patients had a LOS ≥ 5 days (one patient died, one patient had two repairs causing a LOS ≥ 5 days), 104 patients had a LOS < 5 days, but were readmitted ≤ 30 days, and 23 patients had a prolonged hospital stay (≥ 5 days) and were subsequently readmitted within 30 days postoperatively. In 12 patients, the principal reason for prolonged LOS was unclear. Thus, principal reasons were identified through discussions between the investigators.

Repair for incisional and parastomal hernia was significantly more frequent in the study group than in the control group (76% versus 28% and 3% versus 1%, respectively, p ≤ 0.001, Table 1 and Table 2). Poor outcome was found in 26% (95% confidence interval (CI): 23-29) after incisional hernia repair, 30% (95% CI: 14-44%) after parastomal hernia repair and in 3% (95 % CI: 2-4%) after umbilical/epigastric hernia repairs (Table 1 and Table 2). Compared with the control group, the study group was characterised by larger hernia defects (median 8 cm versus 2 cm, p < 0.001) and by a larger proportion of recurrent hernia repairs (21% versus 12%, p < 0.001).

The reasons for poor outcome (prolonged LOS and/or readmission) registered in medical records are shown in Table 3. Major complications (especially wound dehiscence and/or deep infection and cardiovascular events), minor complications (especially seroma formation and infections) and pain were the three main reasons. Among patients with a prolonged LOS, major complications were observed in 30 patients (19%), 43 patients (34%) were readmitted due to major complications and 38 patients (30%) underwent emergency reoperation after the readmission. During the prolonged LOS, 18 patients (12%) underwent emergency reoperation (perforated ulcer (n = 1), wound dehiscence (n = 2), intra-abdominal bleeding (n = 5), ileus (n = 4, one patient died shortly after the operation), anastomotic leak (n = 1) and visceral perforation (n = 5)). Thus, 11 patients (bleeding, anastomotic leak, visceral perforation) had a reoperation as a result of insufficient primary surgery. In 42 (27%) patients, pain with no apparent explanation was the single other important factor causing prolonged LOS. In 14 patients, seroma formation was the primary reason for prolonged LOS. In 19 (12%) repairs, no reason was given in the patient records to explain the long LOS (Table 3).

A total of 43 patients (34%) were readmitted due to major complications and 38 patients (30%) underwent emergency reoperation after readmission (24 deep wound infections, four early ventral hernia recurrences, three perforated ulcers, three visceral lesions, three major bleeding and one ileus). Pain and minor wound infections caused readmissions after 18 (14%) and 14 (11%) repairs, respectively (Table 3).

DISCUSSION

In this case-control study based on nationwide Danish data, we found a poor outcome in 11% of patients undergoing elective repair for a ventral hernia. Important risk factors were repair for incisional, parastomal, and recurrent hernia as well as repair for large hernia defects.

The DVHD is a national database in which all surgeons are obliged to register ventral hernia repairs [3]. The DVHD currently covers 80% of all ventral hernias performed in Denmark [3]. The DNPR is closely related to reimbursement in the Danish health-care system (public and private) and is considered to cover close to 100% of all contacts to the health-care system and surgical procedures in Denmark [5-7]. As in the DVHD, data in the DNPR are based on patients’ social security number and therefore it is possible to combine data from the two databases to obtain data about LOS, readmissions, death, etc. In the present study, we included patients with prolonged LOS and readmission since LOS is probably a valid proxy for morbidity after ventral and inguinal hernia repair [1, 8]. In Denmark, almost all contacts to the health-care system are promoted by the general practitioner [9]. Our results may therefore underestimate the incidence of complications, especially minor wound infections.

Large and complex hernias are a technical challenge to repair and may be beset with a high risk of complications of up to 48% [1, 10, 11]. The presence of complex hernias and technical challenges could in part explain our results, since the control group and study group were comparable in terms of gender and age, but not in terms of type and size of hernia and therefore may reflect findings in the literature [12-15]. Nevertheless, we demonstrated a prolonged LOS even in patients with no obvious complications or clinical complaints as judged by the medical records. It may therefore be argued that optimised perioperative care and fast track regimens are important to improve outcome in patients undergoing ventral hernia repair [16]. Interestingly, our readmission rate and LOS of approximately 3% after umbilical/epigastric hernia repair is comparable with findings from more complex gastrointestinal surgical procedures such as gastric bypass operations [17].

Major and potentially fatal complications caused 19% and 34% of the cases of prolonged LOS or readmission, respectively. Nevertheless, only one patient died within 30 days after elective hernia repair. In 27% of patients with prolonged LOS pain, the only complaint was without relation to hernia size, suggesting that pain reduction is a major topic in the search for better results after ventral hernia surgery [18]. Wound infection was an important cause for readmission after ventral hernia repair. Wound dehiscence and deep wound infections imply serious morbidity and may even lead to recurrent hernia repair [19].

In conclusion, repair for incisional-, parastomal-, recurrent and large hernia defects were important risk factors for a poor outcome. Prolonged LOS and readmission were mainly caused by pain, major complications, wound infections and seroma formation. Thus, there is still plenty of room for improvement in the surgical treatment of complex ventral hernias.

Correspondence: Frederik Helgstrand , Kirurgisk Afdeling, Køge Sygehus, 4600 Køge, Denmark. E-mail: freh@regionsjaelland.dk

Accepted: 28 August 2011

Conflicts of interest:Disclosure forms provided by the authors are available with the full text of this guideline at danmedbul.dk

Acknowledgement: We wish to thank the Danish surgical departments for providing data, and to study research nurse Pernille Strandfelt for her enthusiasm and hard work collecting the data.