An open and closed case: timing of closure following laparostomy

Abstract

Introduction

Laparostomy is important in the management of patients with intra-abdominal gastrointestinal catastrophe or trauma. It carries significant risk and is resource intensive, both in terms of nursing and surgically. The main goal is to achieve prompt myofascial closure (MFC) in order to minimise morbidity and mortality. Early MFC was initially defined as within 2–3 weeks but there is growing evidence that this should be measured in days.

Methods

Retrospective analysis was undertaken of laparostomy cases between 2016 and 2018 at an acute trust and trauma centre serving a population of 500,000. Indication, duration of open abdomen (OA), number of relook procedures and consultant presence were examined to see whether they affected MFC rates, morbidity and mortality.

Results

Overall, 76 laparostomies were performed during the 3-year study period. The most common indication was peritonitis (68.4%). As duration of OA and number of relook procedures increased, the chances of MFC fell significantly. After day 1, MFC rates fell by 20% with each subsequent 24 hours. Leaving the abdomen open primarily at index procedure compared with performing laparostomy following a postoperative complication was associated with significantly higher MFC rates (92.6% vs 68.2%, (p=0.006). The mortality rate was 15.8%.

Conclusions

If the OA is not closed within five days or by the third relook procedure, then achieving MFC is unlikely. Alternative methods should be employed to close the abdomen rather than continuing to take the patient back to theatre for relook laparotomies while increasing the risk of morbidity and mortality. A proactive strategy to forming primary laparostomy at the index procedure has high closure rates.

Introduction

A laparostomy or open abdomen (OA) is a key component of damage control surgery. Indications include trauma, abdominal sepsis, ischaemia where relook is indicated, abdominal compartment syndrome, severe visceral oedema and loss of abdominal wall domain.^1–3 Consequently, patients who require OA are often unstable, high risk and comprise a heterogeneous group. The World Society of Emergency Surgery has produced guidance on OA following a literature review and Delphi consensus process.⁴

During OA, there are several different techniques of temporary abdominal closure (TAC). These include negative pressure wound therapy (NPWT),^5,6 Bogota bag,⁷ Wittmann Patch^® (Starsurgical, Burlington, WI, US),⁸ skin closure only⁹ and fascial traction devices (eg abdominal reapproximation anchor).¹⁰ NPWT appears to be associated with lower mortality than no NPWT.¹¹ There is some evidence that myofascial closure (MFC) rates are higher with NPWT than with other TAC.^11–13 There were initial concerns that NPWT may be a risk factor for enteroatmospheric fistula although Coccolini et al have shown that this is not the case.¹⁴

In our institution, a commercial NPWT with a protective visceral layer (ABThera™ system (KCI, Gatwick, UK) is the agreed method for TAC. This decision was based on the evidence presented: lower mortality, higher fascial closure rates, reduced oedema and contamination, and reduced retraction of the recti muscles as well as the practicality of NPWT in terms of nursing and medical resources for wound management. Figure 1 shows the component parts of NPWT.

Figure 1.

Component parts of negative pressure wound therapy (ABThera™ system)

Optimal timing for re-exploration of a laparostomy patient will depend on the individual case (indication, physiology and local resources). Abdominal compartment syndrome may develop if an OA is closed too promptly. However, during OA, patients are at risk of significant morbidity including intra-abdominal abscess, bleeding, adhesions, enteroatmospheric fistula and mortality.¹⁵ High fluid and nitrogen losses also put patients in a profound catabolic state. Early fascial closure (defined as within 2–3 weeks) is associated with lower mortality and complication rates than delayed closure with mesh.¹⁶ Recent results from an international prospective OA database have shown that this linear correlation exists between complications and duration of OA within days rather than weeks;¹⁷ Miller et al have also found this correlation.¹⁸ Long-term follow-up after damage control laparotomy has also demonstrated better quality of life and functional outcomes in OA patients who underwent closure early versus delayed.¹⁹

Besides minimising morbidity, the main aim of laparostomy is to eventually achieve primary MFC. The unopposed myofascial edges of an OA will retract with time, making primary closure more difficult. World Society of Emergency Surgery guidelines have recommended that the first relook occurs within 24–48 hours of primary surgery.⁴ In the case of the deteriorating patient, a more prompt return to theatre is advocated. Conversely, if a laparostomy is closed too early, then abdominal compartment syndrome may develop.²⁰

Pommerening et al demonstrated that 24 hours after index laparostomy there is a 1.1% decrease in the odds of primary MFC with each subsequent hour’s delay in return to theatre.²¹ They also noted increased intra-abdominal complications with subsequent delays in return to theatre. These American data were exclusively for trauma damage control laparotomy.

In the UK, the most common indication for OA is abdominal sepsis and return to theatre for complications of abdominal surgery.¹⁵ There is a paucity of data on the effect of timing on MFC rates in the non-trauma setting where the pathophysiology is different.²²

Our unit has a lower threshold for laparostomy use than traditional teaching/other centres, which is balanced by early return to theatre to achieve prompt closure. As a major trauma centre, our use of laparostomy has evolved. Following success with NPWT in trauma, we have become more proactive in our use of laparostomy for physiologically unstable patients with intra-abdominal sepsis or ischaemic bowel with potential need for further relook, particularly among those with peritonitis as a result of a postoperative complication, who are at higher risk of abdominal wall dehiscence.

The aim of this study was to examine whether there is a correlation between time to re-exploration of OA, number of relook procedures and primary MFC rates. We hoped to identify whether there is a point at which primary MFC is unlikely to be successful and alternatives must be employed. Alternatives include synthetic or biological mesh, partial fascial closure and planned ventral hernia.²³ We also examined whether there was a difference in outcomes between primary laparostomies at the index procedure and laparostomies required following a postoperative complication.

Methods

Retrospective analysis was undertaken of all laparostomy cases performed between 1 January 2016 and 31 January 2018 at North Bristol NHS Trust, a major trauma centre, busy gastrointestinal unit, and tertiary referral centre for vascular and urological surgery. Cases were identified from a prospectively collected log of all laparostomies, completed in emergency theatre. All case notes were reviewed to evaluate demographics, operative findings, duration of OA, number of relook procedures and outcomes. Patients with an OA who died prior to any method of definitive closure were excluded from the study.

The primary outcome measures were MFC rates, duration of OA in days and number of relook laparotomies. Secondary outcome measures were morbidity: enteroatmospheric fistula, other complications and mortality. Consultant presence at operation and reoperation was also recorded, as was the need for parenteral nutrition. In addition, indication for laparotomy was examined to see whether outcomes differed. Statistical analysis was performed using SPSS^® Statistics version 26 (IBM, New York, US).

Results

A total of 89 laparostomies were performed between 1 January 2016 and 31 December 2018. Thirteen patients were excluded as they died before definitive closure with an OA. After exclusions, laparostomy numbers increased from 21 cases in 2016 to 24 in 2017 and 31 in 2018.

National Emergency Laparotomy Audit (NELA) data over the same period were examined to give a crude denominator of laparostomy use. Over the 3-year study period, 589 laparotomies were recorded in the NELA database. Twenty laparostomy cases did not meet NELA criteria, which would give a crude laparostomy rate of 9.5% (56/589).

Of the 76 patients included in our study, 42 (55.3%) were male. The mean age at presentation was 60.2 years (range: 16–88 years, standard deviation: 17.6 years).

Indications

Peritonitis (sepsis and ischaemia) was the most common indication for laparostomy (n=52, 68.4%). This was followed by trauma (n=10, 13.2%), vascular (n=5, 6.6%) and other causes (n=9, 11.8%). Cases of ischaemic bowel with a high chance of a relook or where a primary anastomosis was high risk were managed with OA. Intra-abdominal sepsis was the most common peritonitis indication (ie 4-quadrant faecal peritonitis secondary to diverticular perforation, sepsis as a result of an anastomotic leak). Other indications for laparostomy included postoperative bleeding requiring packing and diagnostic uncertainty in the face of significant metabolic disturbance necessitating relook on demand.

In 22 patients (28.9%), laparostomy was required as part of treatment of a complication resulting from previous surgery (such as anastomotic leak or bleeding). Twelve cases followed a complication after elective surgery while ten followed emergency surgery.

In 75 cases (98.7%), the ABThera™ dressing (negative pressure wound dressing with protective visceral layer) was used for TAC. One patient (1.3%) had a Bogota bag at the first operation. In all cases, a consultant was present when the decision was made to leave the abdomen open. At the first relook procedure (second operation), consultant surgeon presence was 97.3%.

Outcomes

The primary MFC rate (mass closure) was 85.5% across all indications for laparostomy. MFC rates differed according to indication for laparostomy: 86.5% for peritonitis, 80.0% for trauma, 60.0% for vascular and 100% for ‘other’. The mean time to MFC was 2.1 days, varying from 1.5 days in trauma patients to 3.3 days in vascular patients (Table 1). Time to abdominal closure (irrespective of method) ranged from 1 to 19 days (interquartile range: 2 days).

Table 1.

Myofascial and alternative abdominal closure rates by open abdomen (OA) indication

Indication for OA	n	Myofascial closure				Mesh	Planned ventral hernia	Partial closure
		Yes	No	Rate	Average time
Peritonitis	52	45	7	86.5%	2.1 days	4	2	1
Trauma	10	8	2	80.0%	1.5 days	2	0	0
Vascular	5	3	2	60.0%	3.3 days	2	0	0
Other	9	9	0	100%	1.9 days	0	0	0
Total	76	65	11	85.5%	2.1 days	8	2	1

If patients did not have sutured MFC, alternative methods of abdominal closure were prosthetic material (Vicryl^® [Ethicon, Somerville, NJ, US] or biological mesh; n=8, 10.5%) or planned ventral hernia (n=2, 2.6%). One patient was closed partially and the remainder of the wound left to close via secondary intention.

Almost three-quarters (71%) of laparostomies were performed at the primary operation. The remainder were performed at a subsequent procedure to deal with a postoperative complication. Leaving the abdomen open at the index operation was associated with a significantly higher MFC rate than when laparostomy was performed to deal with a postoperative complication (χ²=6.99, p=0.006) (Table 2).

Table 2.

Comparison of myofascial closure rates for different open abdomen (OA) strategies

Strategy	Myofascial closure	Alternative closure	Total	Myofascial closure rate
Abdomen left open at index procedure	50	4	54	92.6%
OA following postoperative complication	15	7	22	68.2%

The number of relook laparotomies required to achieve abdominal closure varied from one to nine (for one patient). As the number of relook laparotomies increased, MFC rates fell from 96.7% at the first relook to 0% by the fourth relook procedure. Conversely, the need for alternative closure methods (ie mesh/planned ventral hernia) increased from 1.6% at relook 1 to 100% at relook 4 (Fig 2).

Figure 2.

Trends for abdominal closure with each subsequent relook laparotomy after index laparostomy

Similarly, as the duration of OA increased, the chances of achieving MFC fell. If closed within two days of index laparostomy, there was a 100% chance of achieving MFC. After day 3, the chances of achieving MFC fell and after day 6, alternative measures for closure were universally required (Fig 3).

Figure 3.

Myofascial closure rates by duration of open abdomen (OA)

Nutritional support during OA varied. Over a third (39.5%) of patients required total parenteral nutrition (TPN), a similar number (38.1%) required enteral feeding (either nasogastric, nasojejunal or via gastrostomy) and the remainder were fed orally once the abdomen was closed.

Senior consultant surgeon presence was 100% at the decision to perform OA. At the first relook procedure, senior presence was 97.5%. As this was so high, differences in closure rates and mortality were not examined for this variable.

Complications

The mortality rate was 15.8% (n=12). The enterocutaneous fistula rate was 2.6% (n=2) and 1.3% of patients (n=1) developed a vesicocutaneous fistula. Superficial wound dehiscence was the most common postoperative complication (19.7%), usually requiring NPWT therapy. Other complications are recorded in Table 3.

Table 3.

Summary of complications

Complication	n
Superficial wound dehiscence	15 (19.7%)
Mortality	12 (15.8%)
Respiratory (pneumonia, ARDS, lung abscess)	8 (10.5%)
Short bowel syndrome (TPN dependent)	4 (5.3%)
Abdominal collection	3 (3.9%)
Enterocutaneous fistula	2 (2.6%)
Vesicocutaneous fistula	1 (1.3%)
Cardiac (ACS)	1 (1.3%)
Renal failure	1 (1.3%)

ACS = acute coronary syndrome; ARDS = acute respiratory distress syndrome; TPN = total parenteral nutrition

Statistical analysis

An analysis of variance test demonstrated that time to closure (irrespective of method) and numbers of relook procedures were associated with significant differences in MFC rates (both p<0.001). Age (p=0.096), sex (p=0.778) and indication for OA (p=0.416) did not significantly affect MFC rates.

Binary logistic regression of duration of OA and MFC rates was performed. After day 1, for each subsequent 24 hours the abdomen was left open, there was a 20% reduction (Exp(B) 0.208, 95% confidence interval: 0.075–0.571, p=0.002) in the chances of achieving MFC.

Discussion

The indications for OA in our series are consistent with those in the largest previous UK series, reported by Carlson et al.¹⁵ Peritonitis was the most common indication for OA in our cohort. This is often a result of a postoperative complication (22 cases). Over two-thirds (68.4%) of our patients had peritonitis, which is comparable with the findings of Carlson et al (68.9%). It is interesting to note that the lowest rates of MFC were seen in the vascular subgroup (60% vs 80–100%). This group was composed entirely of postoperative aortic aneurysm surgery patients. These patients are more prone to ventral hernia.²⁴ The dysregulation of connective tissue, particularly collagen (which predisposes them to abdominal aortic aneurysm), in these patients may explain the lower rates of primary closure and need for mesh.²⁵

Interestingly, the main difference between our results and those of Carlson et al is in the higher MFC rate for our study (85.5% vs 41.1% for NPWT and 48.4% non-NPWT).¹⁵ Carlson et al reported delayed primary closure rates and hypothesised that NPWT might be used as a method of achieving closure in some hospitals. There was no definition of delayed fascial closure. However, in the literature, 2–3 weeks or longer is classically defined as delayed. This stark difference in MFC rates between two comparable UK datasets is most likely explained by our unit’s strategy for early relook and closure where possible.

This is also evident when comparing our findings with international data for peritonitis. In the international multicentre study by Coccolini et al with 402 patients, the primary MFC rate was 59.3% (vs 86.5% for our cohort) and the mean duration of OA was 5.2 days (vs 2.1 days for our cohort).¹⁷

Our data support the recommendation of the World Society of Emergency Surgery that prompt return to theatre is essential in managing the OA. With each subsequent relook procedure and as duration of OA increases, the chances of achieving primary MFC decrease significantly.

If the abdomen cannot be closed by the time of the third relook, then achieving primary MFC is unlikely. Similarly, if MFC has not been achieved by day 6 of OA, it is extremely doubtful that this will be successful and an alternative approach must be considered. Having a departmental strategy for closure can be helpful. Older studies defined early fascial closure as occurring within 2–3 weeks but it is becoming clear that this should be redefined in terms of a matter of days rather than weeks. Multiple repeated relook procedures are unlikely to achieve MFC and instead increase the risk of serious complications (ie enteroatmospheric fistula). Implementing early abdominal closure as a protocol would save theatre time as it would reduce the need for returns to theatre and probably morbidity.

Our study shows that the rate of MFC falls by 20% per day after day 1. This is consistent with the data of Pommerening et al, who found a reduction of 1.1% per hour (equating to 26.4% per day) after the first 24 hours.²¹ In our cohort, this applied to non-trauma damage control surgery as well.

Reasons why fascial closure may fail are persisting sepsis, loss of abdominal wall domain, fistulation, retraction of rectus sheath and ongoing visceral oedema. Duration of OA and number of relook procedures are of course intrinsically linked to the reasons listed above.

Our finding that a primary index laparostomy is associated with higher MFC rates than a secondary laparostomy for a postoperative complication is logical. This lends support to adopting a proactive rather than a reactive approach to OA formation and has been a driver of our increased use. It may be time to challenge the traditional surgical teaching that a laparostomy should be avoided at all costs given better MFC with modern TAC. This must of course be coupled with early return to theatre. However, there will be instances where this is not possible.

Nutritional support is important in patients requiring laparostomy. Forty-one per cent of our patients required TPN. If possible, enteral feeding is preferred because of the risks associated with central line and cholestatic liver function test derangement with TPN. Early enteral feeding is associated with earlier MFC and lower nosocomial infection rates.^26,27 Nevertheless, enteral feeding should not be used in patients with blind ending/stapled off loops of bowel, which is sometimes the case in OA patients. With reduced time to abdominal closure demonstrated in our data, the use of TPN may be reduced.

Senior consultant surgeon presence is vital to drive decision making for these difficult patients. Consultant presence was almost 100%, which might have contributed to the high MFC rates.

Our mortality rate was 15.8%. This is slightly lower than observed in other OA studies.¹² If our exclusion criteria are ignored, the absolute mortality rate would be 28.1% (25/89 cases). This aligns exactly with what was found by Carlson et al (28.2%).¹⁵

The crude laparostomy rate of 9.5% is based on NELA figures, and accounts for inclusion/exclusion criteria for NELA and our study.²⁸ Our unit’s adjusted 30-day mortality rate for emergency laparotomies (including laparostomies that are eligible under NELA criteria) was 4.8%, making it one of six hospitals in England and Wales with the best outcomes in the 2019 NELA report.²⁹

Study limitations

This was a retrospective study looking at a heterogeneous group of patients. Owing to the small volume of patients, it was not possible to comment on the effect of independent variables on mortality rates. Our study focused primarily on the effects of MFC on the duration of OA, number of relook procedures and strategy for OA. Mortality and fistula rates were secondary outcomes. Patient outcomes such as quality of life and postoperative incisional hernia rates were not examined.

Conclusions

Prompt return to theatre within 24–48 hours for relook laparotomy is essential in OA patients. After day 1, the chance of MFC falls by 20% every 24 hours. If MFC is not achieved by day 5 of OA or the third relook procedure, then an alternative approach to abdominal closure should be considered. Continuing to bring the patient back to theatre for multiple further relook procedures in the hope of achieving MFC is unlikely to be successful.