UPPER GI SURGERY
Effect of handover on the outcomes of small bowel obstruction in an acute care surgery model
- I. Lien; S. W. Wong FRACS, MS; P. Malouf MBBS, FRACS; P. G. Truskett MBBS, FRACS.
An acute care surgery (ACS) model was introduced to manage emergency surgical presentations efficiently. The aim of this study was to evaluate the impact of patient handover in an ACS model on the outcomes of adhesive small bowel obstruction (SBO).
A retrospective study was performed on patients who were admitted with adhesive SBO at Prince of Wales Hospital. The cohort consisted of all patients treated by the ACS team from its introduction in September 2005 to February 2011. Patients in the ACS cohort were divided into two groups: those whose care was handed over to another surgeon and those whose care was not. These groups of patients were compared with a random sample of 50 patients in the pre-ACS period.
In the ACS period, there was no significant difference in complication rates or length of hospital stay in those who were not handed over and those who were. A significantly higher proportion of operations took place during the day for the group who had been handed over (72.7% versus 36.7%; P = 0.005). There were no significant differences in complication rates or length of hospital stay in the pre-ACS and ACS period.
Management under an ACS team does not increase adverse outcomes for adhesive SBO. Patients can be safely handed over within an ACS framework. Other members of the ACS team may help facilitate continuity of care.
The acute care surgery (ACS) model is a consultant surgeon-led model designed to facilitate the emergency workload and to address the unpredictability of theatre access. The model was also developed in response to issues with maintaining general surgical skills. The result of increasing surgical sub-specialization provided disincentives for surgeons to practice outside their area of expertise and perform general surgery on-call duties. Elective lists often conflicted with the emergency schedule procuring a tendency to delay emergency surgery, thus resulting in adverse patient outcomes.
There is now a growing worldwide evidence base of the benefits of ACS models on acute surgery presentations and their management.[4-6] ACS models in the Australasian region have been shown to positively impact on operating theatre utilization and efficiency, staff satisfaction, trainee supervision and opportunities for teaching.[1, 7, 8] Studies assessing patient management and outcomes have established the advantages of an ACS model in managing certain surgical conditions, such as acute appendicitis and acute cholecystitis, with benefits in complication rates, timely surgical intervention and length of hospital stay.[9-11]
The ACS model was introduced at the Prince of Wales Public Hospital (POWH) in September 2005. Surgeons involved in the ACS service relinquish or have limited other commitments during the time period in which they are rostered as the ACS surgeon. The consultant surgeon is on-site and a dedicated theatre is available during standard hours. The surgeon is in control of prioritizing cases and supervising any surgery that is being performed. The ACS surgeon roster is a rotation of 10 general surgeons. Other members of the ACS team include a surgical registrar, resident medical officer, clinical nurse consultant and medical students. In the POWH ACS model, handover is a formalized process between the ACS surgeons between the three rostered periods of the week (Monday 0800 hours–Wednesday 1230 hours; Wednesday 1230 hours–Friday 1800 hours; Friday 1800 hours–Monday 0800 hours). Patients who have not had definitive surgery and may require surgery in the near future can be handed over to the next ACS surgeon. This is at the discretion of the ACS surgeon.
Small bowel obstruction (SBO) is a common cause of emergency hospital admission. It is unknown if the ACS model impacts positively on patients who present with SBO. As SBO is an emergent surgical condition, the provision of an acute care surgical service may impact on the management of these patients. Potentially, ACS may adversely affect patients who present with SBO because they may be handed over from surgeon to surgeon without definitive care. The management of patients with acute appendicitis and cholecystitis is more straightforward, with early surgery being the ideal. Patients who are present with adhesive SBO may not require an operation initially but may require one subsequently because of the development of complications or if the SBO does not resolve with conservative treatment. The aim of this study was to evaluate the impact of patient handover in an ACS model on the outcomes and management of adhesive SBO.
A retrospective study was conducted to patients who were admitted with adhesive SBO at the POWH. The cohort consisted of all patients treated by the ACS team from its introduction in September 2005 to February 2011. Private patients who elected to be treated in the public hospital were included in the study. Private patients who elected to be transferred to a co-located private hospital, patients who were admitted for elective surgery and patients who self-discharged against medical advice were excluded. These groups of patients were compared with a computer-generated random sample of 50 patients treated in the 5-year period prior to the introduction of the ACS. This study was approved by the Human Research Ethics Committee of the South Eastern Sydney and Illawarra Area Health Service.
The following information were recorded from patient's medical records, discharge summaries and operation reports: patient demographics, time and date of presentation and discharge, use of imaging, operative details, and morbidity and mortality. The overall length of hospital stay (emergency department arrival to hospital discharge), time from admission to surgery and post-operative stay in days were recorded.
Patients were grouped according to whether they had been handed over at least once between acute care surgeons and those who had not been handed over. Operation times were divided into two categories defined as day (0800-1700 hours) and after-hours (1700-0800 hours) for further analysis.
Data were recorded using Microsoft Access, imported to Microsoft Excel and analysed using SPSS version 19 (SPSS Inc, Chicago, IL, USA). Chi-square test was used to evaluate categorical data and Mann–Whitney U-tests were used to evaluate continuous data. Results were considered statistically significant if P < 0.05.
There were 171 admissions for adhesive SBO during the ACS period. Of the 171 admissions, 110 were in the cohort who had not been handed over (H0) and 61 in the cohort who had been handed over (H1+). Seven patients were handed over more than once. Demographics were similar between the two groups with respects to age, gender and insurance status (Table 1). Patients who were handed over were significantly more likely to have a computed tomography (CT) scan performed (60% in H0 versus 78.7% in H1+; χ2 = 6.17; 1 degree of freedom (d.f.) P = 0.01) and undergo gastrografin small bowel series (GGSBS) (5.5% in H0 versus 23% in H1+; χ2 = 11.63; 1 d.f. P = 0.001) (Table 2). The median length of stay for those patients who were handed over was less than for those who had not been handed over (5 days versus 6 days). This was not statistically significant (Mann–Whitney U-test, P = 0.19).
Table 1. Patient demographics, investigations and management in the ACS period
|Median age in years (range)||61.5 (19–92)||62 (18–90)||0.87|
|Female||61 (55.5%)||31 (50.8%)||0.56|
|Privately insured||38 (34.5%)||21 (34.4%)||0.95|
|CT scan||66 (60%)||48 (78.7%)||0.013|
|GGSBS||6 (5.5%)||14 (23%)||0.001|
|Surgical intervention||49 (44.5%)||22 (36.1%)||0.28|
Table 2. Patient demographics, investigations and outcomes of patients who had surgery in the ACS period
|Age (median, range) (years)||70 (19–92)||61 (19–87)||0.44|
|Female||31 (63.3%)||17 (77.3%)||0.24|
|Privately insured||18 (36.7%)||9 (40.9%)||0.74|
|CT scan||37 (75.5%)||21 (95.5%)||0.04|
|GGSBS||4 (8.2%)||8 (36.4%)||0.003|
|Number of patients with complications||27 (55.1%)||7 (31.8%)||0.07|
|Mortality||2 (4.1%)||0 (0%)||0.34|
|Median length of stay in days (range)||10 (2–35)||10 (5–20)||0.46|
|Median time to arrival in theatre in hours : minutes||22:28||86:21||<0.001|
|Median post-operative stay in days (range)||7 (1–29)||6 (0–19)||0.33|
|Operation during day hours||18 (36.7%)||16 (72.7%)||0.005|
There were 71 (41.9%) admissions that required operations in the ACS period, with 49 out of 110 (44.5%) in the cohort who had not been handed over and 22 out of 61 (36.1%) in the cohort which were handed over. Demographics were similar between these two groups (Table 2). Of those who underwent surgical intervention, patients handed over were more likely to have a CT performed (75.5% in H0 versus 95.5% in H1+; χ2 = 4.04; 1 d.f. P = 0.04) and undergo GGSBS (8.2% in H0 versus 36.4% in H1+; χ2 = 8.60; 1 d.f. P = 0.003) (Table 2).
The overall proportion of patients who had complications was less in those who had been handed over than those who had not been. However, this was not significant (55.1% (27 out of 49) in H0 versus 31.8% (7 out of 22) in H1+; χ2 = 3.30; 1 d.f. P = 0.07). The most common operation-specific complication was need for bowel resection, and the most common general complication was respiratory (Table 3). There were two mortalities in the group who had not been handed over and no mortality in the group which had been handed over (4.1% in H0 versus 0% in H1+; χ2 = 0.92; 1 d.f. P = 0.34).
Table 3. Complications in the ACS period
|Number of patients with complications||27 (55.1%)||7 (31.8%)||0.07|
|By type of complicationa|
|Confusion/delirium||3 (6.1%)||1 (4.5%)||0.79|
|Recurrent SBO||3 (6.1%)||0 (0%)||0.24|
|Readmission with pain||1 (2.0%)||0 (0%)||0.50|
|Wound infection||3 (6.1%)||0 (0%)||0.24|
|Fluid collection||1 (2.0%)||0 (0%)||0.50|
|Bowel resection||16 (32.7%)||3 (13.6%)||0.09|
|Respiratory||5 (10.2%)||2 (9.1%)||0.88|
|Urinary||3 (6.1%)||0 (0%)||0.24|
|Line complications||2 (4.1%)||0 (0%)||0.34|
|Other infectious||3 (6.1%)||0 (0%)||0.24|
|Pseudomembranous colitis||0 (2.1%)||1 (4.5%)||0.13|
|Cardiac||3 (6.1%)||2 (9.1%)||0.65|
|Significant electrolyte and metabolic complications||3 (6.1%)||1 (4.5%)||0.79|
A significantly higher proportion of operations took place during the day for the group who had been handed over (36.7% in H0 versus 72.7% in H1+; χ2 = 7.88; 1 d.f. P = 0.005) (Table 3). Patients who were handed over had a longer time to arrival in theatre (22 h, 28 min in H0 versus 86 h, 21 min in H1+; Mann–Whitney U-test, P < 0.001). With regard to length of stay, the overall median length of stay (10 days in H0 versus 10 days in H1+; Mann–Whitney U-test, P = 0.46) and median post-operative stay (7 days in H0 versus 6 days in H1+; Mann–Whitney U-test, P = 0.33) were similar in both groups.
Female gender, no previous history of SBO and no history of malignancy were predictive of need for surgical intervention on univariate analysis (Table 4). Abdominal CT scans were performed more often than GGSBS and was more predictive of need for operative intervention (P < 0.001 compared with P = 0.07, respectively).
Table 4. Predictive factors for surgery (for ACS group)
|Median age in years (range)||60 (18–90)||64 (19–92)||0.12|
|Female||44 (44%)||48 (67.6%)||0.002|
|Privately insured||32 (32%)||27 (38%)||0.44|
|History of previous SBO||64 (64%)||15 (21.1%)||<0.001|
|Diabetes||5 (5%)||3 (4.2%)||0.81|
|Cardiovascular co-morbidities||12 (12%)||8 (11.3%)||0.88|
|Malignancy||34 (34%)||12 (16.9%)||0.01|
|CT||56 (56%)||58 (81.7%)||<0.001|
|GGSBS||8 (8%)||12 (16.9%)||0.07|
|CT and GGSBS together||5 (5%)||9 (12.7%)||0.07|
Demographics were similar between the pre-ACS and ACS groups with regard to age, gender and insurance status. Patients in the ACS group were more likely to have CT scans performed (38% in pre-ACS versus 66.7% in ACS; χ2 = 13.27; 1 d.f. P < 0.001). There were similar rates of patients who underwent a GGSBS (16% in pre-ACS versus 11.7% in ACS; χ2 = 0.65; 1 d.f. P = 0.42). There were no significant differences in rates of surgical intervention, bowel resection, morbidity, mortality or median length of hospital stay, time to arrival in theatre or length of post-operative stay (Table 5).
Table 5. Patient demographics, investigations and management of pre-ACS compared with the ACS period
|Median age in years (range)||69 (27–90)||62 (18–92)||0.36|
|Female||25 (50%)||92 (53.8%)||0.64|
|Privately insured||23 (46%)||59 (34.5%)||0.15|
|CT scan||19 (38%)||114 (66.7%)||<0.001|
|GGSBS||8 (16%)||20 (11.7%)||0.42|
|Surgical intervention||19 (38%)||71 (41.5%)||0.66|
|Bowel resection||6 (12%)||19 (11.1%)||0.86|
|Number of patients with complications||14 (28%)||34 (19.9%)||0.22|
|Mortality||1 (2%)||3 (1.8%)||0.91|
|Median length of stay in days (range)||5 (0–59)||5 (0–35)||0.51|
|Median time to arrival in theatre in hours : minutes||55:12||34:47||0.48|
|Median post-operative stay in days (range)||9 (2–58)||6.5 (0–29)||0.10|
The ACS model was introduced to provide an efficient acute surgical service. Surgical outcomes with patient handover in an ACS framework have not previously been examined. There may be potential harm to patients with interruption of the continuity of care. The management and outcomes of adhesive SBO was chosen to be investigated in this study because a conservative initial approach to management is usual, unless there are signs of ischaemia. Immediate operative management of all adhesive SBO cases is supported by the fact that there is an increase in mortality associated with strangulation complications. While other studies have also supported early operative intervention, most surgeons would opt for an initial non-operative period of conservative management if there are no signs of bowel compromise. The optimal time frame for operative management of non-resolving SBO is not well defined. A delay in operative intervention when required has been shown to be associated with an increased morbidity and mortality rate.[13-20]
Handover represents a disruption to continuity of care, which can be detrimental to patient care and outcome. Patient handover is defined as the transfer of patient care responsibility between health-care professionals.[22, 23] In the Prince of Wales ACS model, patients with undefined or incomplete treatment may be handed over to the next rostered ACS surgeon. Clinical handover facilitates continuity of care between surgeons while also helping identify management plans and potential problems. Other members of the ACS team, such as the surgical registrar, who are familiar with the patients from previous days, have an important role in facilitating this continuity of care.[1, 7, 23]
It has been postulated that increased handovers may pose risks for patients through the breakdown of information exchange or due to the emergence of a ‘shift-work mentality’, compromising professional ownership of patients and responsibility of care. Clinical handover to those unfamiliar with a patient has been reported to be associated with an increased risk of preventable adverse events, with poor handover associated with delays in diagnosis, increased rate of complications and increased length of hospital stay.[21, 23, 24]
This study has demonstrated that the management of patients with SBO is not adversely affected with patient handover within the ACS model in our hospital. However, there was no demonstrated improvement in outcomes unlike for acute appendicitis and acute cholecystitis. A comparison between management in the pre-ACS period in which patient handover did not occur and in the ACS period showed similar complication rates and bowel resection rates.
Following the introduction of the ACS model, results show that patients with SBO can be handed over safely with no significant increases in morbidity, mortality or median post-operative stay. Overall, 71 out of 171 (41.5%) patients required surgery, with an overall bowel resection rate of 19 out of 171 (11.1%). The small bowel resection rate was higher in the no handover group, although this was not significant. A greater tendency towards surgical management was not observed in the handover group, which may suggest that surgical intervention was timely. The increased complication rate in patients who had not been handed over may be partially attributable to the severity at presentation, or a delay in initial presentation to hospital.
Analysis of the proportion of surgeries according to time categories showed a significantly greater proportion of patients who were not handed over were operated on after-hours. This was probably related to the urgency of surgery. In contrast, patients who were handed over had a greater proportion of surgeries during the day. This may reflect the availability of time to plan surgery during standard hours for patients who had been admitted for longer. One of the benefits of an ACS model is being able to perform semi-urgent surgery during daylight hours.
There was an increased usage of CT and performance of GGSBS observed in the handed over group. The increased usage may be explained by the extended preoperative time. In our study, CT scan was more predictive of the need for surgical intervention compared with GGSBS, but this may be due to sampling error because CT scans were used much more frequently. The use of CT may provide a clearer clinical picture and prompt early operative intervention when appropriate, thereby reducing the risk of ischaemia and the need for bowel resection.[25, 26] In particular, models have shown mesenteric oedema and lack of small bowel faecalization as CT features, which are significant predictors of the need for operative intervention. The use of a water-soluble contrast agent such as gastrografin has been reported to be effective in both diagnostic and therapeutic capacity in the management of adhesive SBO.[28, 29] While gastrografin has been shown to accelerate the resolution of adhesive SBO, there are conflicting conclusions in the literature regarding its therapeutic role with regard to reducing the need for operative intervention.[30-34] A recent meta-analysis of 14 prospective studies found that administration of a water-soluble contrast agent does significantly reduced the need for surgery.
Some possible explanations exist for our study findings. Traditionally, patient handover has been described as a critical time in patient care due to the potential for incomplete or poor information exchange. Physician-to-physician information relay may be prone to ‘diagnosis momentum’ as erroneous clinical impressions can be carried through.[36, 37] Conversely, it has also been suggested that clinical handover may provide an ‘audit-point’ for patient management and opportunity for collaborative input. Participation of doctors with greater clinical experience may minimize errors in information transfer due to increased acumen in recognizing potential complications. This notion could be reflected in view of consultant handover in the ACS model. Furthermore, such ‘discontinuity of care’ may benefit patients by providing fresh perspective. One study found that the presence of ‘redundant’ systems such as nurses and other residents with knowledge about patients reduced the effect of information errors or omissions in patient handover.
Limitations of this study relate to its retrospective design. Data were restricted to the availability and accuracy of medical records. We had no post-operative information on privately insured patients who arranged for transfer to a co-located private hospital. Thus, these patients were excluded from our study. The numbers in the study may be too small to show significant differences – for example, the median time to arrival to the operating theatre was 20 h longer in the pre-ACS group compared with the ACS group, but this was not significant (P = 0.48).
Several models of ACS delivery exist worldwide. In Australia, individual model variations between hospitals are determined largely by factors, including differences in emergency surgery load and availability of surgeons and other staff. These differences also include variations in ACS surgeon rosters and handover schedule and practices. This study was conducted based on a design at a single hospital, and thus, the study findings may not necessarily extrapolate to a different context.
Despite the increased patient handover between acute care surgeons inherent in the implementation of an ACS model, our study has not shown increased adverse outcomes in the management of patients with SBO. Challenges exist in patient handover as described in the literature which may be applicable to handover between ACS surgeons. However, the role of other ACS team members may help facilitate continuity of care. Overall, our study supports an ACS model for patients with adhesive SBO. Patients can be safely handed over with no increases in morbidity, mortality or length of post-operative stay.