Clinical frailty and its effect on the septuagenarian population after emergency laparotomy

Abstract

Introduction

The UK has an ageing population with an increased prevalence of frailty in the over 70s. Emergency laparotomy for acute intra-abdominal pathology is increasingly offered to this population. This can challenge decision making and information given to patients should not only be based on mortality outcomes but on relative expected quality of life and change to frailty syndromes.

Materials and methods

This was a single site National Emergency Laparotomy Audit (NELA)-based retrospective cohort audit for consecutive cases in the septuagenarian population assessing mortality, length of stay outcome and subjective postoperative functioning. Follow-up was conducted between one and two years postoperatively to determine this.

Results

Some 153 patients were identified throughout the single site NELA database. Median age was 79 years with a ratio of 1.7 men to women. Median rate of all-cause mortality was 35.3% at the median follow-up of 19 months. Median time from admission to death was 120 days. Of those who had died by the time of follow-up, significant preoperative indicators included clinical frailty scale (p < 0.0001), preoperative P-POSSUM (mortality). At follow-up, 35% responded to a quality of life follow-up. This revealed a decline in mid-term physical functioning, lower energy, higher fatigue and reduction in social functioning. There was also an increase in pre- and postoperative clinical frailty scale score.

Conclusion

In the septuagenarian-plus population it is important to consider not only risk stratification with mortality scoring (P-POSSUM or NELA-adjusted risk), but to take into account frailty. Postoperative rehabilitation and careful recovery is paramount. Where possible, during the counselling and consent for emergency laparotomy, significant postoperative long-term deterioration in physical, emotional and social function should be considered.

Introduction

By the end of the 2020s, 22% of the UK population is expected to be over 65 years of age, equating to 15.8 million people;¹ 1/1100 people each year undergoes an emergency laparotomy.² In the UK, the interplay between frailty and emergency general surgery laparotomies remains under significant scrutiny, with the National Emergency Laparotomy Audit (NELA) evaluating outcomes since 2014. Approximately 50% of all emergency laparotomies performed are in the over 75-year-olds. One-quarter of over 85-year-olds can be considered frail. The UK 30-day mortality for emergency laparotomy is 10.6% but in the over 70-year-olds this is almost double at 20%.³

Frailty remains a heterogeneous entity with no distinct difference between normal physiological ageing and frailty. It can be described as a ‘condition or syndrome which results from a multisystem reduction in reserve capacity, to the extent that a number of physiological systems are close to or past the threshold of symptomatic clinical failure’.⁴ This can be explained using a phenotype model, a cumulative deficit model or a comprehensive geriatric assessment.^5–8 With an ageing population in the UK there is increasing interest in frailty in general surgery.^9,10 The Canadian Study of Health and Ageing has found that simplified tools such as a visual scale for categorising frailty can be as effective as formal assessment by a geriatrician.^10,11 It is well established that frailty predicts poorer surgical outcomes in older patients, and those who are classed as frail are more likely to have a longer stay in hospital, increased risk of postoperative complications and discharged to a higher care facility.¹² A 2019 national audit found that one-fifth of patients undergoing emergency laparotomy are frail and have a higher 30-day mortality.¹³

Outcomes of emergency laparotomy in the frail are important; quality of life and change in perceived health arguably more important than longevity. Knowledge of this fact and predictors of outcome will help to guide decision making in this vulnerable population.

Queen Alexandra Hospital is a single-centre large tertiary district general hospital performing around 150 emergency laparotomies per year (inclusive of laparoscopic approaches) with a median age of 68 years.¹⁴ Our preliminary cohort study analyses outcomes in the septuagenarian-plus population who required laparotomy or laparoscopy for an acute surgical abdomen.

Materials and methods

We interrogated the data of all patients over the age of 70 years who underwent emergency laparotomy. These patients were identified over a one-year period from July 2015 to July 2016 using the NELA database. This study was conducted and authorised as a single-centre audit, as all practice was within departmental standards. Patient demographics, disease and procedural factors, clinical frailty scale (CFS) and discharge variables were recorded and analysed. Data were collaborated through retrospective analysis of case notes in order to understand the impact of emergency laparotomy on postoperative quality of life. Patients were reviewed in line with departmental follow-up using a questionnaire (Short Form-36, SF-36, a validated questionnaire to determine the quality of life across eight domains and compare them against national mean scores). The clinical frailty scale (Figure 1) was recorded from the index admission by assessment of activities of daily living, performance status and frailty assessments performed. Data were analysed using univariate and multivariate analysis for disease factors, operative approach and primary outcomes. All non-parametric data were expressed as medians with interquartile ranges, while parametric fields were expressed with means and standard deviations. Primary outcome measure included all-cause mortality at median follow-up, with an aim to capture 18–24-month follow-up data. Secondary outcome measures included length of hospital stay, morbidity, postoperative quality of life and change in the clinical frailty scale. Comparative analysis of those who were alive at follow-up and those who were not was performed to further delineate association of disease, operative and demographic factors thus counselling elderly frail patients prior to emergency general surgery laparotomy.

Figure 1 .

The clinical frailty scale

Results

A total of 153 patients over the age of 70 years (septuagenarian-plus) who underwent emergency laparotomy were identified and included in the study. Case notes were interrogated and cross-referenced against the NELA dataset. Median follow-up was 19 months (range 16–23 months) The median age was 79 years (interquartile range, IQR, 75–84) with a female to male ratio of 1.7 (n = 96 vs n = 57). The 30-day mortality rate was 13% while the 90-day mortality rate was 20%. Mortality at one year was 30.7% (n = 47). This increased to 35.3% (n = 54) at median follow-up of 19 months while the median time to death was 120 days (IQR 17.25–223 day; Table 1). Our Kaplan–Meier survival curve demonstrated a significant change of cumulative proportional survival around to 20 days. This makes the 30-day mortality an important unit figure for longer-term mortality; days 20–100 only demonstrated a further 10% proportional decline (Figure 2).

Table 1 .

Comparative demographics, intraoperative and postoperative factors

	Total (n = 153)	Non-survivors (n = 54)	Survivors (n = 99)	p-value	OR (95% CI)
Age (median + IQR)	79 (75–84)	80.5	79	0.81a
Sex (% male, n)	37.3 (57)	44.4 (24)	31.3 (31)	0.89b
ASA grade 3+ (%, n)	64.1 (98)	66.6 (36)	62.6 (62)	0.61b	1.2 (0.59–2.4)
Malignant pathology (%, n)	29.4 (45)	55.5 (30)	15.2 (15)	0.05b	7 (CI 3.2–15.1)
Contaminated abdomen (%, n)	22.9 (35)	24.1 (13)	22.2 (22)	0.73b	1.18 (0.54–2.60)
Haemorrhage (%, n)	3.3 (5)	5.7 (3)	2.0 (2)	0.26b	2.8 (0.46–17.6)
Perforation (%, n)	10.5 (16)	18.5 (10)	6.0 (6)	0.02b	3.5 (1.2–10.3)
Laparoscopic approach (%, n)	33.3 (51)	33.3 (18)	33.3 (33)	0.56b	1.17 (05.8–2.40)
Clinical frailty score ≥ 5 (%, n)	32.0 (49)	51.9 (28)	21.2 (46)	0.034b	3.2 (1.09–9.61)
Lactate (mean ± SD)	2.35 (1.78)	3.20	1.86	0.001c
Preoperative P-POSSUM mortality (median + IQR)	10.6 (4.1–27.1)	17.2 (6.9–41.95)	10.15 (2.2–14.9)	0.001a
Time to theatre > 48 hours (%, n)	47.7 (73)	50.0 (27)	46.5 (46)	0.68b	1.15 (0.59–2.23)
Length of stay (median days)	16	18.5	15	0.18a
ITU admission required (%, n)	47.7 (73)	43.0 (23)	44.4 (44)	0.73b
Mean length of stay in ITU (%, n)	6.3 (5.5)	7.7 (3.9)	5.5 (7.4)	0.1c

^a Mann Whitey U test

^b Comparison on proportions with 2 × 2 or Chi² Contingency table

^c Student’s t test

ASA, American Society of Anesthesiologists; CI, confidence interval; IQR, interquartile range; ITU, intensive care unit; OR, odds ratio; SD, standard deviation from the mean

Figure 2 .

Kaplan–Meier survival curve

Some 94.8% of patients were admitted through the emergency pathway; the further 5.2%, (n = 8) were elective admissions with unexpected postoperative complications requiring laparotomy (two postoperative anastomotic leaks, one post operative chylous leak, one peritoneal catheter complication, one post cystectomy bleed, one wound dehiscence, one refashioning of the ileostomy, and one elective admission who required a Hartmann's procedure for diverticular abscess). The overall median length of stay in hospital was 16 days (IQR 10–30 days).

Patient demographics and outcomes were analysed between non-survivors and survivors (Table 1); 47.7% of patients in the overall cohort had their emergency laparotomy after 48 hours. This was equivalent in both groups (p = 0.6). There was no significant difference in the average age of the two groups (80 years vs 79 years, p = 0.98), sex (44.4% male vs 31.3% male, p = 0.89) or those admitted as an emergency (94.7% and 93.9%). Those deceased at the median follow-up had a significantly higher initial venous lactate (3.2 vs 1.86, p = 0.001) and a significantly higher P-POSSUM score (17.2 vs 10.1, p = 0.001).

There was also a trend to a longer hospital stay (18.5 days vs 15 days, p = 0.18) and longer intensive care stay (7.7 days vs 5.5 days, p = 0.11) in the non-survivors, although this was not significant. In our cohort, the operative approach or duration of greater than 48 hours to surgery did not affect whether the patients were likely to be alive at 19 months. Malignant pathology appears to be a significant factor in predicting outcome after major laparotomy in septuagenarians. Over 50% of patients deceased at 19 months post-emergency laparotomy had underlying malignant pathology (55.5% vs 15.2%, p < 0.05). The majority of those alive had benign disease (84%), with obstruction being the most common indication for theatre. Perforation also appears to be a negative prognostic indicator, with a significantly higher mortality in this group (n = 18.5 vs n = 6, p = 0.02). Most patients sustained no gross contamination; this was equal with no difference across both groups.

A significantly lower preoperative median CFS was observed in those still alive at follow-up compared with those that died within the follow-up period (CFS score 4 vs 9, p < 0.0001) Following surgery there was an observed deterioration in the CFS from an average score of 4 to 5 (now mildly frail). Using a multivariate logistic regression analysis, the clinical frailty scale was highly significant for mid-term mortality, particularly when adjusted with P-Possum (p < 0.05). A CFS score over 5 had the best-fit receiver operating characteristic (ROC) curve; this demonstrated an area under the ROC curve of 0.86. This area can be increased to 0.89 when the CFS is adjusted for P-POSSUM using a multivariate analysis and logistic regression modelling (Figure 3).

Figure 3 .

Receiver operating characteristic curve for clinical frailty score 5 or above compared with midterm mortality (area under the curve 0.86)

Of the total cohort, 47.7% (n = 73) were discharged back home with no significant extra support. A further 9.8% were discharged with a hospital, at home, or physiotherapy support package and 3.3% were discharged with a full package of care. Thus, overall, 39% of septuagenarians undergoing emergency laparotomy were not discharged back to their normal residence: 12.4% (n = 19) were discharged to rehabilitation or intermediate care facilities, 4.6% were discharged to a residential home, 8.5% were discharged to a nursing home hospice or palliative care hospital. The final 13.7% (n = 21) were not discharged prior to death.

The SF-36 questionnaire was compared against a mean score of the general population across eight different domains and whether or not there was a significant change in health. These included physical functioning, role limitations due to physical health, role limitations due to emotional health, energy and fatigue, emotional wellbeing, social functioning, pain and general health. Of those who underwent telephone follow-up, 35% of patients felt that they were better than prior to surgery, 57% felt worse overall than prior to surgery and 8% thought that there had been no overall change. The cohort of patients who survived to a mean follow-up of 19 months on average reported significantly lower physical functioning, lower energy and more fatigue than the general population. Social functioning was also significantly reduced. A higher CFS was observed in those who did not survive to the mid-term review (Table 2). In the remaining cohort of those who survived, there was a decline in frailty status moving from vulnerable to a mildly frail status. Patients were generally more likely to perceive themselves as the same or worse subjectively. Subjective assessment of overall change in health over the year was not significant when compared with the general population score. There was a statistical decline across physical function and wellbeing and energy levels.

Table 2 .

Short Form-36 score cohort compared with mean population score

Quality of life domains	Our study mean score	General population mean score	p-value	95% Confidence interval
Physical functioning	47.8	70.61	0.001	34.52–61.06
Role limitations (physical)	49	52.97	0.64	31.3–66.7
Energy and fatigue	40.3	52.15	0.029	29.7–50.9
Pain	60.7	70.77	0.16	46.3–75.1
General health	49.6	56.99	0.20	38.1–61.1
Emotional wellbeing	59.7	70.38	0.097	46.9–72.5
Role limitation (emotional)	69.2	65.78	0.70	50.99–87.40
Social functioning	62.7	78.77	0.031	48.3–77.14
General health change	52.9	59.14	0.334	39.8–66.0

Discussion

The impact of frailty in emergency laparotomy is a key area of expansion in emergency general surgery. A 2019 UK national audit reported a similar 30-day mortality of 14.6% and 90-day mortality of 19.5%,¹³ with increased 90-day mortality in the frailer subgroups. Our preliminary cohort has a longer duration of follow-up and shows a mortality of 35.5% at 19 months. We also add evidence for progressive frailty pre- and post-surgery and the potential negative effect on quality of life. In our increasingly ageing population, this is key information and helps patient and surgeon decision making. It highlights the importance of the patient’s wishes, counselling against futile surgery, diligent postoperative management and avoidance of ‘failure to rescue’. Furthermore, comprehensive care of the elderly support and considerate counselling of outcomes are essential in those who do undergo emergency laparotomy.¹⁵

There are multiple ways of assessing and defining frailty, for example with physiological deficits such as sarcopenia and osteopenia or traditional risk tools.^{10, 11} One study assessing P-POSSUM with osteopenia and mortality in geriatric emergency laparotomy found that by accounting for osteopenia one could improve the area under the ROC curve from 0.59 to 0.83.¹⁶ There is a higher incidence of intra-abdominal pathology potentially requiring emergency laparotomy, a higher burden of postoperative complications and more complex social and care challenges in the older, frailer cohorts.¹⁵

In community-dwelling individuals, those classified as pre-frail and frail already have lower quality of life score (SF-36) in their physical and mental wellbeing domains.¹⁷ Although frail patients with a clinical frailty scale of 5 or above have an increased risk of short and mid-term mortality, it should not be the only factor considered when offering surgery.¹⁸ Further considerations with increased likelihood of mortality are those over 90 years, septic shock, American Society of Anesthesiologists grade V (a patient with severe systemic disease that is a constant threat to life).¹⁹ This is not, however, completely unexpected to a surgeon with some experience.

In those over 75 years presenting with acute abdominal catastrophes requiring laparotomy (open or laparoscopic approaches), CFS and P-POSSUM can help to determine intermediate mortality. In those who survive, we can establish preliminary quality of life changes. Care must be taken particularly when counselling patients and counselling family. Our study shows a decline in clinical frailty after emergency laparotomy, with a high proportion of patients not being discharged back to their usual place of residence. Equally, knowledge of the initial insulting pathology is important, as worse outcomes are seen with malignancy and perforation.

This is a preliminary study supporting the need for further studies to more accurately evaluate the difference in progression of frailty and deterioration in quality of life. Continued follow-up at three years would be of interest to delineate whether further recovery improves this frailty status.

Conclusion

Emergency laparotomies for palliative and malignant conditions should be thoroughly counselled in expectations and facilitating early and palliative discharge should be considered from the offset. We need further information alongside the continuing auditing for emergency laparotomies, for patient outcomes and their link to frailty. In our ageing demographic, this should be a key area in which to undertake further study and audit.