In the present review, we found that postoperative complications were common after elective colorectal procedures, occurring in 35.1% of the patients who underwent abdominal surgery. To our knowledge, this is the first Swedish study to demonstrate the direct correlation between surgical outcome and hospital costs in a consecutive series of elective colorectal operations. The most frequent complications recorded as surgical site infection, wound disruption, suture line failure as well as peri-operative haemorrhage might be considered provider related and potentially avoidable.
Elective colorectal surgery has been shown to have a low mortality rate but a high morbidity rate ranging from 9% to 31%. Notably, use of fast-track protocols in colorectal cases has been found to improve the outcome, seen as a 50% reduction in the number of patients with complications after surgery [12–14]. Despite a target length of stay of 3 to 5 days in fast-track protocols, such as the one applied currently, patients without complications in our series were hospitalized for a median of 9 days. Even though standard postoperative care routines were established, achieving continuous optimization of protocol adherence proved to be a challenge. Clearly, further prospective studies are needed to identify the rate of compliance with multimodal rehabilitation in the present unit.
Surgical site infection (SSI) and anastomotic leak have been described as the most common adverse events in several investigations [15–17]. Internationally, SSI is considered to be a major indicator of the quality of surgical care, and it has been estimated that 40% to 60% of SSIs are preventable . The rates of superficial and deep wound infection noted in the present study are similar to data published by Hawn et al. and Tang et al., and compare favourably to values reported in several earlier publications [21–23]. Male sex, obesity, specific surgeons or hospitals, ostomy and total/subtotal colectomy were identified as risk factors for surgical site infection after colorectal surgery [22–25]. The circumscript number of cases present in our cohort did not allow subgroup analysis with regard to any technical aspects associated with SSI.
The incidence of anastomotic leak in elective cases has been found to vary between 1.4% and 7.9%, with the higher value noted in cases involving rectal procedures using an extraperitoneal suture line. Large series (> 800 patients) reported overall anastomotic leak rate under 4% [17, 20, 24, 26].
Moreover, the risk of anastomotic leak has been observed to be increased by long operation time, low level of colorectal anastomosis, and comorbid conditions [27, 28]. The absence of consensus regarding the definition or the diagnosis of anastomotic leak makes it difficult to compare the rates of anastomotic failure published in the literature. Here, we considered dehiscence of the suture line in a blind colon or rectal segment to be similar to an anastomotic leak in terms of morbidity, and we included patients with disruption of the distal stump after Hartmann’s procedure in this group. Both clinical and radiologic anastomotic leaks were registered. Yet, it is probable that the number of radiologic anastomotic leaks was underestimated as asymptomatic patients did not undergo routine radiologic investigation post-operatively.
The incidence of midline wound disruption in our investigation was high compared to rates of 0–2.7% after elective procedures in contemporary studies [5, 17, 29]. Suture technique monitored by the ratio of suture length to wound length (4–6/1) has proven to be the most important factor affecting this variability , which indicates that midline dehiscence could easily be rectified by dissemination of appropriate knowledge and adherence to evidence-based techniques.
Our results showed that age did not increase the risk of complications, which is contrary to previous findings and could be attributed to the fact that only elective cases were included [5, 17]. Large studies have shown that > 10% weight loss prior to surgery and overweight constituted independent risk factors for morbidity after colorectal surgery [16, 30]. As in the present report, several investigations [16, 25, 30, 31] recognised that BMI above 25 and obesity were associated with increased risk for surgical complications. Complex procedures were also correlated with higher complication risks in our study, as has been reported by other researchers . The rate of re-operation after colorectal surgery was deemed to be an independent factor for surgical quality. In the present review this factor compared favourably with contemporary studies .
The attending surgeon was also found to be an independent risk factor for complications in a fixed-variable statistical model. We regarded this model most appropriate, because the surgeons in the present cases were not chosen at random, but were instead assigned to specific operating lists according to their knowledge, training, and experience. In agreement with our findings, these aspects of surgeons have been found to have an impact on surgical outcome in numerous other studies [20, 34, 35]. Drolet et al. observed decreased morbidity and mortality in patients who underwent elective colorectal procedures performed by experienced surgeons, and Hubner and colleagues  noted that the surgeon constituted an independent risk factor for SSI after colon surgery. Surgical skills were hard to assess as the large number of assessment methods indicated . Yet, these skills were recognized as essential in the occurrence of complications . We identified a large difference in the odds ratio for complication between individuals which indicated a distinct disparity within this group of surgeons.
Major part of the adverse events (wound infection, anastomotic failure, wound dehiscence, and perioperative bleeding) observed in our patients was avoidable, which is in accordance with contemporary studies [4, 8]. These reports also found technical error to be the most common error type in surgery. Presumably, some patients will always develop such complications, but, for any given patient, that should not be an expected outcome. It is also obvious, that, once identified, avoidable adverse outcomes, should be readily amenable to rectification.
Our analysis showed that postoperative complications were associated with a substantial and statistically significant increase in total hospital costs, even after adjustment for type of surgery and comorbid conditions. The relative increase ranged from 207% for any complication to 380% for suture line failures, and, concomitantly, length of hospital stay was significantly prolonged.
Nursing and postoperative or intensive care costs made the main contribution in this context, whereas the contribution from surgical operation costs was smaller than we had expected. Expenditures related to reoperations during index hospitalization,per se, could not be retrieved separately from the database we used. Costs were further increased by readmissions, even if the currently observed readmission rate of 7.4% compares favorably with results obtained in earlier studies [39, 40].
These results were consistent with findings by other investigators who have demonstrated a linear correlation between surgical outcomes and accrued hospital costs. Robust evidence showed that programs for continuous quality improvement in surgical care, based on measurement and monitoring of outcome- and process- based quality indicators, were effective in reducing post-operative morbidity, mortality and hospital costs [6, 41–43]. In their analysis of 676 US hospitals, Fry et al. , indexed outcomes and efficiency to a reference set of cost effective hospitals and were able to demonstrate that inefficiency was the major contributor to excess cost. Colorectal procedures in the study population had a complication rate of 7, 7% in index and 16,3% in inefficient sites which resulted in 100% cost increase per incurred complication, a level that is significantly lower than the corresponding rates noted in the present review. As we identified that adverse outcomes resulted in very high average costs per event, it is clear that important cost savings could be achieved by reducing the occurrence of complications.
The present observational cohort study had a number of limitations. The data were derived retrospectively from consecutive patients treated at a single tertiary care unit during 2010, and hence the results might not be applicable to other institutions or other time periods. In an attempt to eliminate the inaccuracy of administrative data in identifying complications, we conducted a systematic review of the surgical records of all in- and out-patients in the index population and collected information on operations from the surgery-specific database. Unfortunately, the surgical unit did not have a structured post-discharge surveillance program for SSIs, and thus there was probably an underestimation bias. Outcomes for discharged patients were evaluated 60 days postoperatively, and therefore long-term outcomes such as incisional hernias and aberrations in urinary and intestinal functions were not registered. The hospital internal accounting database was based on estimates correlated to diagnosis and did not permit detailed estimates of expenditure such as hourly physician cost or medication and medical supplies cost per patient. Nevertheless, the deduction that postoperative complications are independently related to high hospital costs and prolonged hospital stay remains true.
In conclusion, potentially avoidable complications following elective colorectal surgery were frequent in the tertiary care unit we studied, and this resulted in a considerable increase in hospital costs and resource expenditures. Focus on the improvement of surgical outcomes contributes substantially to better quality of care and patient safety and facilitates effective cost containment and resource utilization in healthcare.