The impact of surgical site infection—a cost analysis

Data presented in this manuscript are based on the RECIPE trial. The RECIPE trial was a single-center, randomized controlled, prospective trial with two treatment groups comparing intraoperative subcutaneous wound irrigation with 250 ml of 0.04% polyhexanide solution (Serasept2®, Serag-Wiessner, Naila, Germany) to 250 ml of 0.9% saline in elective open or laparoscopically assisted abdominal surgery. The primary outcome data — rate of SSI 30 days postoperatively — was published previously [6]. Here, we report a post hoc analysis of this prospectively collected cohort analyzing the impact of SSI on inpatient costs and 30-day postoperative outcome.

The RECIPE trial was an investigator-initiated Medicinal Products Act trial. The study protocol was approved by the State Office for Health and Social Affairs Berlin (Study Protocol code RECIPE2014; EudraCT number: 2014–001,551-22) and the Federal Institute for Drugs and Medical Devices. The trial was conducted in accordance with the ethical principles of the Declaration of Helsinki and the principles of Good Clinical Practice (ICH-GCP E6) [7]. This trial is registered at http://​www.​ClinicalTrials.​gov (ID: NCT04055233).

Patients 18 years or older, capable to give informed consent, and undergoing elective open or laparoscopically assisted abdominal surgery were eligible to participate. Patients were randomly assigned either to the control group with 0.9% saline or to the experimental group with 0.04% polyhexanide solution [6].

In the RECIPE trial, SSI was defined according to the criteria by the Centers for Disease Control and Prevention within 30 days postoperatively (Table 1) [6]. Examination of the wound at least every second day, regularly dressing changes, and decision, if SSI was present, were performed on the surgical ward by attending surgeons. To evaluate SSIs after discharge, an interview by telephone was conducted at least 30 days postoperatively. Patients were asked whether the wound was closed or secondary healing and whether inpatient or outpatient medical treatment of the SSI was required. In case of uncertain wound sites, an appointment in our outpatient department was arranged for the patient. We classified SSI in superficial, deep, and organ/space SSI according to CDC (Table 1) [6]. An intraabdominal abscess was counted as organ/space SSI (grade 3) when it occurred in close proximity to the surgical incision.

The endpoints of this post hoc analysis included the overall inpatient costs. The cost analysis referred to inpatient treatment. Data on costs were collected with the help of the tool of our corporate controlling team. The calculated costs of our study population only refer to the department for general and visceral surgery of Charité Berlin, Campus Benjamin Franklin, where the study was conducted. The calculation of costs is based on the official guidelines of the German Institute for remuneration in hospitals (InEK) for the calculation of diagnosis-related groups in Germany. The German Institute for remuneration in hospitals (InEK) sets these guidelines to ensure that the calculation of costs is comparable between all participating hospitals in Germany [8]. The tool of our corporate controlling team is a graphic display which outlines the calculation of costs and the comparison of proceeds and costs. This tool calculated the exact inpatient costs based on the case number of every single patient. These included overall costs of hospital treatment, surgery costs, costs on surgical ward, medication costs, laboratory costs, and costs of diagnostic procedures. Overall inpatient costs consisted of all categories that are depicted in Table 2. Surgery costs were mainly referring to cutting-suture time and time of anesthesia. Surgery costs also involved costs for revision surgery. Costs for medical material and drugs used in the operation room are included. In case of readmission within 30 days postoperatively, costs were added to the primary case number. In the case of reoperation surgery, costs of the second operation were added to the primary case number. Patients were interviewed by telephone at 30 days or greater postoperatively. Satisfaction with wound cosmesis was graded on a numeric analog scale (NAS) from 0–3 (not content) over 4–7 (moderately content) to 8–10 (very content). The parameter pain was assessed in the area of the scar or secondarily healing wound 30 days postoperatively. The pain was recorded on NAS ranging from 0 (no pain) to 10 (strongest pain imaginable).

Categorical variables of the primary outcome were analyzed with cross-tabulation and chi-square tests. For quantitative outcomes, statistical group comparisons were performed using the t-test for independent samples. Due to the skewed distribution of some of the quantitative variables, group differences were analyzed by Mann–Whitney U test or Kruskal–Wallis test. Additional parameters were depicted according to their scale and distribution with absolute and relative frequencies for categorical parameters and mean, standard deviation (SD), median, and interquartile range (IQR) for quantitative parameters; p-values ≤ 0.05 were considered as statistically significant. Statistical analysis was carried out using IBM SPSS Statistics 25® (IBM, Armonk, NY, USA).

In the RECIPE trial, 456 patients were recruited and 393 (86.2%) completed the follow-up 30 days postoperatively [6]. Patients of this post hoc analysis had an average age of 58.3 ± 16.2 years. Overall, there were 154 (39.2%) female and 239 (60.8%) male patients. There was neither a difference in the duration of surgery nor in the organ site of the conducted operation between patients with and without SSI. Right- or left-sided hemicolectomy was performed in 23.9%, low anterior rectum resection in 13.8%, and colectomy or proctocolectomy in 13.0% (Table 3).

The overall inpatient costs for patients with SSI were 16.685 €; 19.703 USD median (IQR 21.638 €; 25.552 USD) compared to 11.235 €; 13.267 USD median (IQR 11.564 €; 13.656 USD) for patients without SSI (p < 0.001). Surgery costs differed in patients with SSI (6.664 €; 7.870 USD) to patients without SSI (5.040 €; 5.952 USD) (p = 0.001). Regarding costs on surgical ward, the occurrence of SSI was more expensive (8.404 €; 9.924 USD vs. 4.690 €; 5.538 USD; p < 0.001). The same refers to medication costs (663 €; 783 USD vs. 306 € (361 USD; p < 0.001) and laboratory costs (965 €; 1.140 USD vs. 592 €; 699 USD; p = 0.005) (Table 4). Additional costs directly caused by SSI were identified by excluding patients with severe postoperative complications (grades 3–5 according to Clavien–Dindo) in a supplementary analysis. Overall inpatient costs for patients with SSI without severe complication were 11.908 €; 14.038 USD median (IQR 4.933 €; 5.816 USD) vs. 9.354 €; 11.027 USD median (IQR 6.306 €; 7.434 USD) patients without SSI (p = 0.006). Costs on the surgical ward were 5.715 €; 6.737 USD median (IQR 4.568 €; 5.385 USD) for patients with SSI vs. 3.707 €; 4.370 USD median (IQR 2.988 €; 3.523 USD) for patients without SSI (p = 0.001).

Overall inpatient costs for patients who were irrigated with saline in the RECIPE trial were 12.056 €; 14.237 USD median (IQR 13.612 €; 16.074 USD) vs. 12.793 €; 15.107 USD median (IQR 12.085 €; 14.271 USD) in the polyhexanide group (p = 0.52). Surgery costs amounted to 5.156 €; 6.089 USD) in the saline and 5.596 € (6.608 USD) in the polyhexanide group (p = 0.53).

Surgery costs varied depending on the performed procedure. For example, costs for esophagus resection were 11.586 €; 13.682 USD median, for pancreas resection 8.473 €; 10.006 USD median, and for right- or left-sided hemicolectomy 4.055 €; 4.789 USD median; p < 0.001 (Table 5).

Overall, we recorded 111 (28.2%) SSIs. Sixty-six (59.5%) of SSI were detected during inpatient treatment and 45 (40.5%) occurred within 30 days postoperatively. Data about the satisfaction with the cosmetic result 30 days postoperatively were documented in 318 patients. Eighteen patients with SSI (16.2%) and twelve patients without SSI (4.3%) were not content (p < 0.001) (Table 6).

The total length of hospital stay including readmission was 18 days median (patients with SSI) compared to 12 days median (patients without SSI) (p < 0.001) [6]. In the group with SSI, postoperative pain at the scar 30 days postoperatively was rated 1.3 ± 1.7 SD on average on NAS compared to 0.6 ± 1.4 SD in the group without SSI (p < 0.001) (6). Patients without SSI complained less often about postoperative pain at the scar than patients with SSI: 71.3% vs. 41.4% had no pain (p < 0.001) [6].