A national UK audit of suprapubic catheter insertion practice and rate of bowel injury with comparison to a systematic review and meta‐analysis of available research

Limited data exist on the risks of complications associated with a suprapubic catheter (SPC) insertion. Bowel injury (BI) is a well‐recognized albeit uncommon complication. Guidelines on the insertion of SPC have been developed by the British Association of Urological Surgeons, but there remains little evidence regarding the incidence of this complication. This study uses contemporary UK data to assess the incidence of SPC insertion and the rate of BI and compares to a meta‐analysis of available papers.

insertions recorded in the UK between April 2017 and March 2018. 1 Twenty-five percent of SPC insertions are performed in an emergency setting, 2 usually when urethral catheterization has failed or is contraindicated. Emergency SPC procedures are often performed on the ward or in the emergency department, commonly using a closed (or blind) technique. Elective SPC insertion is more often performed in theatre with cystoscopy (+/− ultrasound) guidance for patients requiring long-term bladder drainage. Some cases are performed using an open cystotomy approach, for example in patients with previous abdominal surgery where there is an increased risk of anterior abdominal wall bowel adhesions. 3 Limited high-quality data exist on the risks of complications associated with SPC insertion. Bowel injury (BI) is a well-recognized albeit uncommon complication and usually requires emergency surgical treatment. As highlighted in the National Patient Safety Agency (NPSA) rapid response report in 2009, 4 this is a life-threatening complication, a fact that should be taken into consideration during the consent process. 5 Practice guidelines on the insertion of SPC that aim to reduce the risk of BI were developed by the British Association of Urological Surgeons (BAUS), 5 but there remains little contemporary evidence regarding the incidence of this complication or how it may be avoided.
The objectives of this study are to assess the current national incidence of SPC insertion and the rate of BI from a contemporary UK dataset. We also performed a systematic review and meta-analysis of available research on BI rates from SPC insertion.

| MATERIALS AND METHODS
A search of the National Hospital Episodes Statistics (HES) data was performed on all SPC insertions between April 2015 and October 2016 for operating procedure codes, Code M38.2 (cystostomy and insertion of a suprapubic tube into bladder). 1 Patient age, 30-day readmission rate, 30-day mortality rate, and catheter specific complication rates were also collected. Cause of death was not collected.
To estimate the BI rate, we searched for patients who had undergone any laparotomy or bowel operation (Table 1) within 30 days of SPC insertion. As HES is unable to determine the reason for these subsequent procedures, individual NHS hospital Trusts were contacted and an anonymized questionnaire completed to ascertain whether the indication for subsequent bowel operation was an SPC-related BI (Appendix 1 and 2). Unless it was related to BI the reason for laparotomy was not collected.
Accuracy of coding for SPC insertion was analyzed by performing a retrospective subset analysis of the last 100 SPC insertions at a single NHS hospital trust site.
Systematic review and meta-analysis were performed with a PubMed search to identify papers reporting on SPC related BI. Search terms used were "suprapubic catheter" and "bowel injury" or "complications." We excluded any expert opinion, case studies, individual reviews, and included only those that reported BI rate. This left eight papers. Year of publication, location of publication, case number reported, and rate of BI were collected. database recorded SPC insertions, 7347 (64%) were elective, 3866 (34%) emergency, and 260 (2%) unknown. The single-site subset analysis showed that 99 of 100 (99%) SPC insertions were coded correctly. The one case that was not coded correctly was a patient who had an SPC change rather than insertion. This suggests the HES coding search parameters used were appropriate and validates the coding and capture process.

| RESULTS
The all-cause 30-day readmission rate was 17%, which includes admissions for reasons unrelated to the SPC. There were 618 30-day readmissions coded for catheter specific complications (5%), most commonly a mechanical catheter problem (59%).
The total all-cause mortality rate within 30 days of SPC insertion was 167 (1.5%). The cause of mortality as documented on the patient death certificate was not collected. Of these 22 (13%) were elective insertions, 142 (85%) emergency insertions, and 1 (0.6%) unknown.
One hundred forty-one patients underwent laparotomy surgery within 30 days of an SPC insertion in 80 different NHS trusts, (some at same time as SPC insertion). The operation performed at laparotomy for each of these cases is tabulated below ( Table 1). Reason for laparotomy was not collected.
Responses from the BAUS UK hospital survey were received for 114 (81%) of the cases. Of the reported laparotomies, there was one recorded BI associated with SPC insertion. In this case, the SPC was inserted by a subconsultant grade urologist in the operating theatre using a trocar technique. No preceding risk factors for BI were identified (previously reported risk factors include body mass index [BMI] >35, previous lower abdominal surgery, small capacity nondistended bladder). The injury was identified within 24 hours of SPC insertion and was repaired with subtotal colectomy. The patient made a full recovery.
The BI rate in this large, contemporary UK series was, therefore, one in 11 473. Data from previous studies are summarised in Table 2.

| META-ANALYSIS
Analysis of data before the 2009 NPSA report revealed two studies with datasets ranging from 185 to 219, which reported an incidence of BI of 2.4% to 2.7% and a mortality rate of 1.8%. 6,7 Metanalysis of these data shows a BI rate of 10/404, 2.4%.
An overall metanalysis of the available data shows a BI rate of 11/1490, 0.7% (Table 2).

| DISCUSSION
In 2009 the NPSA, a forerunner of NHS Improvement, issued a rapid response report on minimizing risks of catheter insertion. This stated that from September 2005 to June 2009 there were reports of three deaths and seven cases of severe harm relating to the insertion of SPCs. Recommendations from the NPSA report include the use of ultrasound (US) guidance when the bladder cannot be palpated despite adequate filling (>300 mL). They recommend the use of open SPC insertion in cases of previous lower abdominal surgery or patients whom the bladder cannot be adequately distended (<30 mL). They acknowledge that there is no sufficient evidence base to support the use of US during SPC insertion, however, and when utilized that it should only be performed by trained individuals. 14 Following the NPSA report, in 2011 BAUS published guidelines for SPC insertion, recommending that the US may be used as an adjunct to SPC insertion but only by a practitioner with appropriate training and experience. BAUS recommend the use of US in patients for whom the bladder cannot be palpated due to obesity. In contrast to the NPSA report, BAUS recommend that in patients with previous lower abdominal surgery or a bladder that will not distend open SPC insertion or US guidance can be used. They also comment that there is a conspicuous lack of research data and a need for further information on this. 5 The disparity between BI rates reported before and after the NPSA report in 2009 (2.4% and 0.09%, respectively) may be explained in terms of changes in technique, availability of newer Seldinger devices for SPC insertion (such as the Mediplus S-Cath System), or patient selection factors following advice from the NPSA and BAUS. This contemporary series of 11 473 patients undergoing SPC insertion identified only one BI. This study analysis would suggest that the risk of BI is significantly lower than historical data suggest and is more in line with smaller recent reports.
It is likely that the majority of bowel injuries will have been identified in this study, yet some may have been missed through poor coding. Analysis of coding at a single Trust indicates that coding was correct in 99% of patients, although it is accepted that there would be variation in coding amongst different trusts. Our HES search might also have missed cases where BI was managed nonoperatively, but it is anticipated that successful conservative treatment of BI is rare. The range of post-SPC procedures identified for further investigation was comprehensive and deliberating overinclusive, and it is unlikely that any BI-related operations were missed. We also recognize that some cases of BI may have been missed due to a delayed presentation. A small proportion of patients only present with BI at the time of the first SPC exchange. 15,16 The majority of bowel injuries would be expected to present within 30 days of insertion, however. It is therefore unlikely that the BI risk is substantially higher than stated.
Our data include 727 patients aged under 20-year old. It is likely that the risk factors for BI are very different in this group. This group did not undergo any postoperative bowel procedures and no bowel injuries were identified.
We acknowledge limitations to this study. Data are missing for 27 of the total 141 laparotomies performed (19%) and the BI rate may, therefore, be underestimated. However, even if it was assumed that all such cases had a BI secondary to SPC insertion (which is unlikely) then the BI rate would still be in line with recent smaller studies at 0.24% (28/11473).
Although we have not been able to investigate possible risk factors for BI, it is accepted that the risk of BI is higher in certain patient populations (eg, high BMI and previous lower abdominal surgery) and US guidance is recommended in such cases to identify the bladder. 4 However, the matter of exactly how US guidance should be used has not been addressed and there is a variation in practice. Many clinicians advocate the use of intraoperative US to identify bowel loops during needle insertion. This is challenging, even for an experienced operator, and must be performed by a clinician with expertise in USguided interventional procedures.
The thirty-day mortality rate in this series is 1.5%. The mortality rate is not caused specific to SPC insertion and may reflect the age and comorbidities of the groups of patients requiring SPC insertion, some of whom may be receiving an SPC as part of ongoing palliative care. We have no access to cause of death though HES data, and a detailed analysis of the cause of death is beyond the scope of this study.
This study suggests that the risk of BI after SPC insertion is lower than previously reported. To proceed with SPC insertion without US guidance in patients in the absence of risk factors for BI would appear to be safe, and patients can be reassured that the risk of BI is less than 0.25% as in the BAUS patient information leaflet. 17  On the ward or ED (A&E) In the operating theatre In the radiology department What type of procedure was performed?

| CONCLUSION
Blind trocar insertion (with or without cystoscopy) USS-guided trocar insertion (with or without cystoscopy) Insertion using Hargroveʼs (or similar) urethral sound Open insertion Which of the following risk factors for BI were present? BMI >35 Previous lower abdominal open surgery Small capacity or nondistended bladder When was the BI identified?
During the SPC insertion procedure Within 24 hours of the SPC insertion procedure During the same admission After discharge from hospital What was the outcome of the BI?
Repair of damage with full recovery Repair with permanent adverse effects Death What were the outcomes of the trusts safety incident investigation?