Background
The ageing population, along with advances in anaesthetic and surgical techniques, will lead to an increasing number of frail older patients undergoing surgical interventions. Preoperative frailty is associated with increased risk of adverse outcomes. This was objectively quantified in the first study of frailty and surgical outcomes by Mackary and colleagues in 2010, which demonstrated the association of preoperative frailty with increased risk of postoperative complications, increased length of stay (LOS), and discharge to institutional care [
1]. Since then, there has been a surge in literature on the impact of frailty on perioperative outcomes [
2]. Not only is frailty consistently associated with risk of major morbidity, mortality and readmissions [
2‐
5], it is also associated with new patient-reported disability [
6], institutional care, functional decline, and lower quality of life post-surgery [
2,
4].
Despite the strong evidence that preoperative frailty in surgical patients results in poor postoperative outcomes, there is limited evidence to date supporting interventions in frail surgical patients. A 2017 systematic review by McIsaac et al. [
7] found that few interventions improved outcomes in this patient population. Five of the 11 included studies tested multicomponent interventions and these studies failed to consistently demonstrate improvements in outcomes and most were at high risk of bias [
7]. We aimed to systematically review the evidence for multicomponent perioperative interventions in frail patients that has emerged since McIsaac et al.’s [
7] review.
Methods
Protocol and registration
The protocol for this systematic review was registered with PROSPERO (CRD42021282937) and conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines [
8].
Search strategy
We searched PUBMED, EMBASE, Cochrane Central Register of Controlled Trials, CINAHL online databases, with publication dates from January 1, 2016 to October 20, 2021, with updated searches on August 27, 2022 and March 29, 2023. The search terms combined Medical Subject Headings (MeSH) and free text words (See Supplementary for full search strategy). Publications were limited to English language. Additional publications were identified by searching reference lists of included papers.
Study selection
Two reviewers, VK and HT, in the initial database search and, VK and NR or VK and EG in the updated searches, independently screened titles, abstracts and full texts. Reasons for exclusion were documented. Discrepancies on whether a study met inclusion criteria were resolved by discussion and consensus.
Inclusion and exclusion criteria are outlined in Table
1. Included studies were randomised controlled trials or quasi-experimental studies of perioperative multicomponent interventions in frail surgical patients aged 18 years and over. Studies were to use a valid frailty measurement tool. This was defined as a composite measure of deficits in two or more health domains. Studies using a single domain measure (such as a physical performance test) were therefore excluded.
Table 1
Inclusion and exclusion criteria
Population | Aged ≥ 18 years Underwent surgery (all surgical settings including elective/emergency/major or minor surgeries or procedures) Used a valid frailty measure and majority of the sample classified as frail. | Aged < 18 years Did not undergo surgery Did not use a valid frailty measure. Used a valid frailty measure but majority of the sample not classified as frail or data for frail group unable to be extracted. |
Intervention | Perioperative multicomponent intervention | Interventions targeting a single health domain Interventions forming part of established standard of care protocols, such as ERAS. |
Comparator | Standard/usual care Alternative intervention (superiority trial) | |
Outcome | Examined relationship between intervention/comparator and one or more outcome(s) | |
Study design | Randomised controlled trials, quasi-experimental studies | Observational studies |
Publication Criteria | Published and “in press” articles reporting original research results | Conference abstract only, reviews, book chapters, editorials, theses Full text not available |
Language | Studies written in English | |
Perioperative multicomponent interventions were defined as interventions directly related to the patient having or having had surgery that addressed at least two health domains and/or involved two or more healthcare disciplines. Studies evaluating established standard of care protocols only, such as Enhanced Recovery After Surgery (ERAS) protocols, were excluded. There were no inclusion or exclusion criteria relating to the type of study outcomes.
Data extraction and analysis
Data extraction was conducted by VK and verified by NR and EG using pre-specified data fields as agreed upon by all reviewers. Data included country, study design, sample size and characteristics (age, sex), type of surgery, frailty measure, intervention details (description, timing during perioperative period, setting), and overall study outcomes. Due to the heterogeneity of study designs, interventions and outcomes, a formal meta-analysis was not possible. A narrative synthesis of the results was conducted.
Assessment of risk of bias
Risk of bias assessments were conducted for all studies using the National Institute of Health Quality Assessment Tool [
9] by VK or NR and verified by EG.
Discussion
Our systematic review of literature published since 2016 identified five studies of perioperative multicomponent interventions in frail patients undergoing elective oncologic gastrointestinal surgery. There were two RCTs and three quasi-experimental study designs and overall, the quality of the evidence was deemed to be fair. The studies did not consistently demonstrate improvements in outcomes. Reductions in postoperative complications, mortality, length of stay and functional deterioration were reported yet methodological issues and substantial heterogeneity of the interventions precludes drawing clear conclusions regarding the optimal model of care.
In 2017, McIsaac et al. [
7] also found that studies of multicomponent interventions did not consistently demonstrate improvements in outcomes. They attributed this, in part, to poor adherence and protocol implementation issues [
7]. Certainly, there is evidence for a dose-response relationship between ERAS protocol adherence and clinical outcomes after major colorectal surgery [
15] and it is probably reasonable to expect a similar effect with other perioperative interventions. Protocol adherence was identified as an issue in two studies [
10,
11] included in our review. The interventions in these two studies were embedded within a well-established ERAS program and, interestingly, the authors speculated that the study interventions may have had a limited effect, especially with respect to surgical outcomes, given that other aspects of perioperative care were optimised [
10,
11].
The sample sizes of included studies were modest – only one study was adequately powered for the primary outcomes (and found no significant difference between the intervention and control groups for any outcome) [
10]. The quasi-experimental studies [
12‐
14] were retrospective and there were important differences in baseline characteristics in both quasi-experimental [
12,
13] and RCT studies [
11]. Statistical analyses were largely unadjusted, failing to account for potential confounding factors. For example, a statistically significant difference in mortality rates in intervention and control groups in Mazzola et al.’s [
12] study may have been confounded by differences in the rates of pancreatic cancer, a malignancy associated with an extremely poor prognosis [
16].
There was substantial variability among the interventions tested and, as such, it is difficult to ascertain which elements are the key ingredients for an effective intervention in this setting and patient population. All studies addressed physical activity and nutrition, which is in keeping with ERAS guidelines for elective colorectal surgical patients [
17]. Nutrition and physical activity interventions, addressing protein/caloric supplementation and resistance-based training, respectively, are also recommended for the management of frailty more generally [
18]. Health domains known to be relevant to the care of frail adults, such as social support, was not addressed by the interventions in the included studies and only one study [
11] included a review of medical co-morbidity and medications in their intervention.
Comprehensive Geriatric Assessment (CGA) is a comprehensive evaluation by a medical specialist with expertise in geriatric medicine to identify and address medical, social and functional needs, optimise medication prescribing, and engage a multidisciplinary team to assist frail patients to attain goals [
19]. It is, by definition, a multicomponent intervention. Recommended by Best Practice Guidelines as the approach to managing frailty in
all patients [
20], CGA has been shown to increase the likelihood of frail inpatients being alive and in their own homes at follow-up [
21]. The study by Ommundsen et al. [
11], which we included in our systematic review, described an intervention including a geriatric assessment and tailored management plan. This intervention appears to align with the principles of CGA; however, the authors of the study noted that there was minimal access to multidisciplinary allied health input and the time between assessment and surgery was very short (median = 6 days). The intervention did not appear to reduce the rates of traditional adverse surgical outcomes in this relatively small study, which is consistent with meta-analysis of data from studies of preoperative CGA in elective non-cardiac high-risk surgery [
22]. Even so, preoperative CGA is recommended in recent clinical practice guidelines for the perioperative care of frail people undergoing surgery [
23].
Three studies included in our review were ‘prehabilitation’ studies implementing interventions between five days and four weeks prior to surgery [
10,
12,
13]. Prehabilitation is designed to improve an individual’s resilience prior to elective surgery [
24]. The evidence suggests that preoperative interventions need to be implemented a reasonably long time, at least four weeks, prior to surgery in order to build physiological reserve [
25]. However, this is not always feasible. In cancer surgery, for example, delays in treatment can result in poor oncological outcomes. Neoadjuvant therapy increases the time from diagnosis until surgery [
26] and is associated with a decrease in overall physical fitness, which has been associated with worse outcomes after surgery [
27,
28]. Variations in the timing of intervention likely contributes to variability of the study results reported here. Notably, due to there being “no evidence that prehabilitation programmes improve postoperative outcomes for older patients or those living with frailty” (p. 3), current guidelines advocate the use of CGA [
23].
The intervention evaluated by Suzuki and colleagues [
14] primarily occurred in the postoperative period. While it was not a CGA, the intervention addressed multiple health domains with the support of a multidisciplinary team of medical and allied health professionals. Although it was a small study, it demonstrated significantly lower complications and dependence in the intervention group. Compared with the other studies included in this systematic review, the results of Suzuki et al.’s [
14] study may be more generalisable to
emergent surgical populations who are able to receive postoperative (and not preoperative) interventions. CGA conducted postoperatively in hip fracture patients, for example, has been shown to reduce the risk of mortality, readmission and new institutionalisation [
29].
Heterogeneity of outcome measures was identified by McIsaac and colleagues [
7] as a key issue in their systematic review. In our review, effectiveness was primarily measured using a variety of traditional surgical outcomes. It is possible, however, that perioperative interventions in frail surgical patients will have minor effects on traditional outcomes and major effects on patient-centred outcomes such as functional decline, quality of life and discharge disposition. In our systematic review, four studies reported on patient-centred outcomes. None of the studies examined effects of intervention on delirium or cognitive decline, which along with functional decline, are increasingly prioritised by older patients and are of critical importance to informed surgical decision-making [
30]. Furthermore, despite measuring frailty at baseline, none of the studies examined changes in frailty status following multicomponent interventions as an outcome measure. Patient-centred outcomes are not commonly evaluated in clinical trials of frail patients in hospital [
31]. The evidence suggests that many patients with severe illness would not elect for life-sustaining treatment if the burden of treatment was high or if treatment resulted in significant cognitive and functional impairment [
32]. We agree with McIsaac and colleagues that ascertaining what outcomes are most important to frail patients and the people who care for them is necessary to inform future clinical trials. This is a focus of ongoing work by our group.
Strengths and limitations
This systematic review used a comprehensive search strategy with defined inclusion and exclusion criteria, which was broad enough to encompass all types of surgery, in elective and emergent settings, yet narrow enough to permit a synthesis of evidence relating to a particular population group and type of intervention. This review therefore provided important insights into the current state of evidence of the effectiveness of multicomponent perioperative interventions in frail surgical patients.
There are limitations to this study. Despite the broad search strategy, all included studies were conducted in elective oncologic gastrointestinal surgery populations, limiting generalisability of results. The small number of included studies may reflect our protocol’s requirement that a validated measure of frailty be used and that the majority of the study sample be classified as frail. This resulted in exclusion of studies of multicomponent interventions in what may be generally accepted to be frail patient populations, such as orthogeriatric models of care in hip fracture. Nevertheless, as it is well-understood that frail surgical patients are clinically different to non-frail surgical patients and clinical practice guidelines emphasise the importance of using validated tools to diagnose frailty in surgical patients, only including studies that used a validated frailty measure ensured that the evidence presented here is clinically relevant.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.