In this study of patients undergoing major elective surgery, we found that sepsis after surgery was not associated with the CCS. However, a higher CCS was associated with an increased level three length of stay.
The strengths of this study were that it was conducted prospectively and all eligible patients were included. No changes were made to the anesthetic given for the surgery or the management of the patients after surgery, so the conditions in the study reflect those that exist in the hospital environment. In this hospital, it is standard practice for patients who are undergoing thoracic, upper and lower gastrointestinal major surgery to be offered an epidural. Epidural analgesia may reduce the incidence of SIRS through an effect on the surgical stress response after surgery,[19
] although the effect may be limited and the clinical significance is unclear.[21
] Despite widespread use of epidural analgesia in this study, 59 (58.4%) patients developed the SIRS postoperatively. This is a lower incidence than that reported in post-surgical intensive care unit patients (93%).[22
] However, not all of the patients in the current study were cared for in critical care. The incidence of SIRS in the current study is similar to the number of patients developing SIRS (excluding the first postoperative day) in a similar study where no epidural analgesia was provided (46.2%).[8
The rate of sepsis after surgery reported in different reports varies, and can be as high as 40.2%.[23
] A study on the association of postoperative sepsis with the Charlson score reported an incidence of 20%.[8
] Our incidence of sepsis (27%) is comparable with these figures.
There are a number of limitations in this study. The study population included 101 patients, but the number who developed sepsis was relatively small, limiting the ability to detect an association between patient factors previously found to be associated with sepsis. However, our study had a similar number of patients compared to a previous study[8
] that had reported an association between a high Charlson score and postoperative sepsis. Secondly, the study was conducted in one institution, which is a university teaching hospital and tertiary referral center for hepatobiliary surgery. Caution should be used before applying our results to other hospitals where the case mix and postoperative care may be different. The specialized nature of study site and its highly selected population limits generalizability of the study.
Thirdly, as we did not evaluate white cell count daily unless a full blood count had been requested by the treating surgeons, patients who would have met the SIRS criteria with a high white cell count may have been classified as not having SIRS (false negatives). In turn, this could have resulted in a number of patients with sepsis not being identified as such.
In a previous study by Mokart and colleagues, a high CCS was associated with severe sepsis.[8
] In this study, we found no association between CCS and sepsis. Apart from the different endpoints used in the two studies (sepsis in this study, severe sepsis in the former study), there were differences in the patient populations enrolled into the two studies, which may explain this discrepancy. The patients in the current study had a lower average CCS (a median of 2 in this study compared to a mean of 5 in the Mokart study).[8
] Other studies looking at the effect of a higher CCS on outcome have found an adverse effect only when the CCS is higher than 5,[15
] so it may be that in our subgroup of patients with sepsis, the level of co-morbidity was not high enough to have any effect. Secondly, only patients who had operations longer than 5 hours were included in the Mokart study,[8
] whereas we included all patients who were planned to have major elective surgery on an “intention-to-treat” basis, even if the actual surgical time was short. Finally, all patients in the previous study were admitted to a critical care area after surgery. The present study adds to the debate as to whether the Charlson score will always reliably predict the incidence of postoperative sepsis.
We did not find that male gender or age, as reported previously,[5
] were significantly associated with sepsis. The previously reported difference in the occurrence of sepsis between males and females may be due to hormonal differences in estrogen levels.[25
] However, the exact reason is unclear.
It has been reported that blood transfusion during surgery is associated with infections after surgery.[4
] In this study, we found an association between blood transfusion and sepsis on univariate analysis that became nonsignificant on multivariate analysis [Tables and ]. This suggests that the apparent association of blood transfusion with postoperative sepsis may be explained by confounding with longer operating times. Blood transfusions during surgery may confound with prolonged, difficult surgery; intraoperative hypotension; urinary catheterization; or tracheal intubation - factors which could also be associated with the development of sepsis after surgery.[3
We found that the length of operation, on multivariate analysis, may be associated with sepsis postoperatively, with a P
value of 0.054. This suggests that after technically demanding surgery or when complications occur, sepsis is more likely. This is consistent with previous studies that demonstrated an increased risk of wound infection as the duration of surgery increased.[27
Unplanned admissions consumed a greater number of critical care days than planned admissions and the length of stay was significantly longer after unplanned admission (median 6.5 days) than after planned admission (median 2 days, P
= 0.016). Unplanned admission to intensive care after surgery is associated with a longer hospital length of stay and may be an indicator of patient safety.[6
] We found that patient co-morbidity, as reflected in a higher CCS, was also associated with a longer level three unit stay, as has previously demonstrated.[30
It may also be argued that the presence of co-morbidities, such as malignancy, in our study would make patients more prone to developing sepsis after surgery. We agree that this is a distinct possibility. However, there has never been any study that has compared the incidence of postoperative sepsis in cancer with non-cancer patients.[8
] Additionally, one would assume that an increased length of stay would also correlate with the presence of co-morbidities. However, this is not inevitable since it has been shown that using a postoperative care pathway in identified patients can significantly reduce length of stay.[31
In this study, we found that the duration of surgery, but not patient co-morbidity as assessed by the CCS, may predict the incidence of sepsis postoperatively. The CCS could be used as a guide to predict the likely consumption of critical care resources by elective surgical patients, as a higher CCS was associated with a longer level three unit stay in this study.
The targeting of critical care resources using either preoperative optimization[32
] or postoperatively using early goal-directed therapy[33
] to high-risk surgical populations has been shown to reduce hospital length of stay and improve outcomes. There was a demonstrated decrease in the incidence of infective complications by 40% using preoperative optimization, which consisted of increasing oxygen delivery to 600 ml/min/m2
by using intravenous colloid and dopexamine. It has been estimated that in the UK, if about 500 patients annually could be treated with this perioperative protocol,[32
] up to 20 lives could be saved. It was also estimated that because of the projected reduction in hospital stay, there should be an annual cost saving of at least £2,000,000 (approx. INR 140,000,000). With increased availability of advanced non-invasive hemodynamic monitoring, this could be carried out outside of a critical care unit, leading to further cost savings.[34
Additionally, by identifying patients at risk and allocating personnel resources, such as a rapid response Medical Emergency Team (MET), the relative risk of postoperative sepsis can be reduced. The introduction of such a team, which included the duty intensive care doctor and a designated intensive care nurse with an emergency pack with drugs and equipment needed for resuscitation and tracheal intubation, led to a relative risk reduction in postoperative sepsis (74.3%; P
= 0.0044) as well a reduced hospital length of stay [23.8–19.8 days (P