Evaluating the implementation process of a multicomponent perioperative intervention for prevention of postoperative functional decline in geriatric patients: a qualitative study

Abstract

Objectives

Geriatric patients are at increased risk of developing postoperative neurocognitive disorders, including delirium. Existing evidence-based perioperative interventions need to be implemented into routine care to improve postoperative outcomes. In this qualitative interview study, we wanted to collect stakeholder experiences to understand the implementation process of a multi-component intervention to prospectively facilitate future implementation.

Design

Descriptive qualitative evaluation research.

Setting

Single-centre at a German major urban academic hospital.

Participants

22 interviews were conducted with n = 7 geriatric patients after surgery who had received a comprehensive geriatric assessment and an individualised perioperative multi-component intervention, and n = 15 healthcare professionals, including nurses, physicians and medical assistants working in the perioperative care.

Interventions

Semi-structured interviews were conducted, addressing the implementation procedure of the multicomponent intervention and the experience with it within the routine setting.

Main outcome measures

The implementation outcomes were adoption, acceptance, appropriateness, feasibility and sustainability.

Results

Transcribed audio recordings were analysed with directed content analysis. Most intervention components could be adopted during the pilot trial. Implementation barriers were identified. Limited resources and logistic constraints threatened feasibility and sustainability. Acceptance of patients and healthcare providers regarding an intervention depended on its perceived appropriateness, which varied per intervention component, workspace and duration of the implementation.

Conclusions

We were able to replicate and extend previous findings on the implementation of improved perioperative care. To facilitate the implementation success and motivation to implement evidence-based measures, resource allocation needs to be adjusted and standard operational procedures, as well as the cross-sectional collaboration, must be simplified.

Trial registration number

NCT03325413.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Implementation evaluation of evidence-based perioperative multicomponent intervention under real-life circumstances for various surgeries.

• Qualitative method enabled in-depth data collection of the implementation process and its components.

• Capturing multidisciplinary experience from patients and various healthcare professionals from different workspaces and backgrounds.

• Risk of selection bias due to the small number and non-random selection of interview partners.

• No long-term data were collected, so it is unclear how many of the components, or to what extent, could be sustainably implemented into clinical routine.

Background

In 2019, 7.1 million out of 18.8 million inpatients in Germany underwent surgery. More than half of these patients (3.6 million) were 60 years of age or above.¹ Older adults are particularly at risk of developing postoperative complications (POCs), including postoperative neurocognitive disorders (pNCD), including delirium.² These adverse outcomes can last from brief moments to medium and long term, impacting patient-relevant parameters such as functional status, coping with daily activities and quality of life.³

Besides age, several intrapersonal and procedural factors are known to influence the risk of postoperative functional decline. Intrapersonal risks include malnutrition, frailty, pre-existing cognitive malperformance, poor physical fitness, anaemia, potentially inappropriate or multiple medications, as well as preoperative anxiety and stress.⁴,⁶ Procedural risks include prolonged preoperative fasting, disorientation, intraoperative blood loss, hypothermia, deep anaesthesia, opioids and anticholinergic medication.⁷,⁹

Several evidence-based recommendations and medical guidelines for the perioperative management of geriatric patients aim to reduce POCs and enhance recovery.¹⁰,¹³ Various targeted preoperative prehabilitation programmes have been tested in clinical trials with different foci, entailing comprehensive screening for risk factors, tailored nutritional supplementation, respiratory and physical exercise, the reduction of anaemia and/or addressing psychological stress and disorientation.¹⁴,¹⁸

However, literature on the implementation of tailored and comprehensive perioperative interventions to reduce pNCD in the routine care rather than controlled settings is sparse.¹⁹,²¹ While several qualitative evaluations of comprehensive geriatric assessment (CGA) and/or prehabilitation bundles have been published recently, these are mostly tailored to a specific patient cohort and/or type of surgery, and it remains to be seen how their findings can be transferred to other settings.²²,²⁵ There is also little information available on individual intervention components operating immediately before or intra-surgery. Implementing multicomponent interventions in hospital care affects many different stakeholders. Patients, different workspaces and professions are involved in different aspects, acting independently and possibly unaware of other interventions.²⁶ The experiences of different stakeholders should be explored to understand which and how interventions can be implemented in the routine perioperative care of geriatric patients. This knowledge could help to successfully reduce the pNCD risk across areas and medical centres.

A sound assessment of the implementation process and barriers and facilitators is achieved by using robustly conceptualised implementation outcomes. In their Implementation Outcomes Framework (IOF), Proctor and colleagues proposed a heuristic taxonomy, identifying several implementation outcomes as key targets of change.²⁷

In a recent monocentric, prospective, non-randomised, controlled pilot study, we implemented a multicomponent perioperative intervention within routine care for general elective surgery. The anaesthesia consultation was updated to screen for risk factors and introduce individualised perioperative procedures at multiple time points and sites to reduce the risk of postoperative functional decline, see Olotu et al.^{28 29} In this qualitative study, we aimed to evaluate the implementation from the perspective of healthcare professionals (HCPs) and patients to contribute to the improvement of perioperative care of geriatric patients. The attributed research question ‘Can an evidence-based, perioperative, multicomponent intervention to reduce postoperative functional decline be implemented into the routine care, and what are barriers and facilitators of a successful implementation?’ was answered by using the outcomes feasibility, adoption, appropriateness, acceptance and sustainability selected from the IOF.²⁷

Methods

Design and setting

As part of a prospective controlled pilot study, we monitored the implementation process of a preoperative CGA followed by a tailored preoperative and intraoperative multi-component intervention. The study took place within the routine care of the University Medical Centre Hamburg-Eppendorf. To evaluate the implementation, we conducted semi-structured interviews with patients and HCPs. The investigated multicomponent intervention addressed different risks and potential mitigation strategies at different points in the perioperative process. A detailed description of the intervention components can be found in Olotu et al.²⁹ We used the Consolidated criteria for Reporting Qualitative research (COREQ)³⁰ and template for intervention description and replication (TIDieR) reporting guidelines (onlinesupplemental appendices A C).³¹

Pilot study

To implement and explore the effects of the interventions, 278 patients aged 65 and above with impending surgery were recruited in the pilot study. They had been consecutively allotted to three groups: routine care (the control condition), implementation phase and intervention phase (the experimental condition). The implementation phase was used for the gradual implementation of all components, leaving space for adoption, tailoring and modifications as necessary. After the implementation of all intervention components, the recruitment of the intervention phase was initiated, see Olotu et al.²⁸ Some interventions were indicated for all patients. Other interventions were applied to patients with specific risk factors; for example, iron supplementation was administered only when iron deficiency or anaemia was present (table 1). Effectiveness of the intervention was estimated by Instrumental Activities of Daily Living, the primary clinical outcome, and secondary outcomes, including POCs, cognitive performance, quality of life and length of hospital stay 1 and 6 months postoperatively.³² Fidelity and penetration of the intervention were assessed by differences in implementation rates between groups.

Table 1. Intervention components in temporal order, indication and involved departments.

	Intervention	Indications/risk factor	Involved departments	Modifications
Preoperative interventions	Iron supplementation (1× intravenous administration)	Iron deficiency anaemia	PreMed/surgery department	Indication: surgery with estimated bleeding risk ≥10%; recommendations to discuss with general physician if thought applicable
	Supplementation of folate and cobalamin	Deficiency states	PreMed/surgery department	Not implemented (ingredient of the high-protein supplementation)
	Medication optimisation	Multi-medication	PreMed/pharmacy	Not implemented, recommendations to discuss with general physician if thought applicable
		Potentially inadequate medication
	High-protein nutritional supplement 2× daily until surgery	Malnutrition	PreMed/home	Indication: always
		Frailty
	Daily physical exercise Daily breathing exercise	Frailty	PreMed/home	Indication: always Modification: no on-site training demonstration, but brochure hand-out
		Reduced grip strength
		Reduced mobility
		Reduced endurance
	Involving confidants in preoperative patient education	When applicable	PreMed	Not implemented (structural confines, few confidants)
	Delirium prevention education	Always	PreMed
	Carbohydrate loading	Always	Inpatient ward/home
Intraoperative interventions	Availability of glasses, hearing devices and dentures in the OR	When applicable	Inpatient ward/surgery department	Transitional permission to bring personal items in OR area from the department of hygiene and department of QM. Installation of lockable cabinets in the OR area to safely store items during surgery
	Patient warming	Always	OR-holding and PACU	Officially implemented already; consistently established here
	Use of perioperative pain catheters whenever possible	Always	OR/OR-holding and PACU/inpatient ward
	EEG-guided anaesthetic administration	Always	OR
	Regional anaesthesia	When possible and appropriate	OR
	Avoidance of potentially inadequate medication* in the perioperative phase	Always	OR/OR-holding and PACU/inpatient ward

^*Benzodiazepines, anticholinergics, central alpha-agonists, application of short-acting muscle relaxants.

EEG, electroencephalography; OR, operating room; PACU, post-anaesthesia care unit of the surgery department; PreMed, anaesthesia outpatient clinic; QM, quality management.

Cohort and inclusion criteria

Following the maximum-variation principle of convenience sampling, participants were selected based on the interventions they gave (HCP) or received (patients). That is, the sample included at least one patient with each orientation aid, respectively, at least one frail patient and one patient flagged for potential multi-medication. Patients underwent surgery in at least two different operating theatre (OR) sections and were sent from six inpatient wards; hence, we aimed to represent different surgeries and conditions (table 2, online supplemental table S2). At least three HCPs were interviewed per workspace in the perioperative timeline: the pre-anaesthesia consultation at the anaesthesia outpatient clinic; inpatient care at the various inpatient wards; and preoperative preparation and surveillance in the OR (tables 1 and 3). Nine patient interviews were planned. This number was chosen in an attempt to balance investigating different conceptual outcome categories with the ethical concern of including many patients.^{33 34} Content saturation was assumed.

Table 2. Demographic characteristics of participants.

Patients
Age, M (SD), years	74.14 (6.52)
Female sex, N (%)	3 (42.9)
Treating department, N (%)
Urology	1 (14.3)
Gynaecology	2 (28.6)
General medicine	1 (14.3)
Casualty surgery	3 (42.9)
Healthcare professionals
Age, M (SD), years	45 (12.92)
Female sex, N (%)	11 (73.3)
Profession, N (%)
Technical assistant/medical student	5 (33.3)
Nurse/nursing assistant	5 (33.3)
Physician/resident	5 (33.3)
Workspace*
Pre-anaesthesia consultation	8 (33.3)
Preoperative and postoperative inpatient care	5 (20.8)
Preoperative preparation and surgery	11 (45.8)
Involvement in the implementation process	7 (46.7)
Working experience, M (SD), years	18.47 (14.28)

^*Multiple responses possible.

M, mean.

Inclusion criteria for patients were participation in the intervention group and reception of at least one tailored intervention for a specific indication (table 1). The interviewer knew the name, age and tailored intervention(s) for patient participants and the ward affiliation for HCP participants. Participants were informed of the content, rationale and use of the interview material and gave written consent to participate. 15 interprofessional HCPs (physicians, assistants, ward nurses, OR nurses, surgery and anaesthesia outpatient coordinators) involved in and/or affected by the implementation were interviewed.

Procedures and instruments

Interviews were conducted face-to-face in an office in the medical centre, or for patients who were still treated in inpatient care, in their room with no other persons present. Six of the HCP interviews took place during the implementation phase. These interviews were used to gain insight into the implementation process and to allow for any necessary adjustments to it. Participants in the implementation phase interviews were all directly involved in the implementation process. All other interviews took place after the implementation was realised. Patients were interviewed a maximum of 6 months after their elective surgery, at their last follow-up assessment. The interviewer (psychologist with a master’s degree, female) was informed about the procedures and had attended monthly project update meetings, but was not directly involved in the implementation. A semi-structured interview guide was developed in advance by the evaluation team, clustered according to predefined outcomes, and was pilot tested with a study associate (online supplemental appendix B).

Outcomes

The outcomes for the qualitative evaluation were feasibility, adoption, appropriateness, acceptance and sustainability, selected from Proctor et al’s working taxonomy of outcomes for implementation research (table 3).²⁷ Interviewee-driven topics were welcomed and not suppressed.³⁵

Table 3. Implementation outcomes as defined by Proctor et al²⁷.

Outcome	Definition
Adoption	Intention and commitment to the implementation, including any necessary adaptations to fit the setting
Acceptance	Satisfaction of those affected by the implementation
Appropriateness	Perceived relevance and compatibility with the setting
Feasibility	Degree of successful implementation
Sustainability	Level of preservation in the intended setting

Analysis

The interviews were audio-recorded. Field notes were made during the interview. Basic transcription rules were followed.³⁶ Colloquial language and dialect were approximated to standard language, and non-verbal cues were not annotated. The transcribed data were pseudonymised, that is, all names and attributable content were removed. Transcripts were not returned to participants, as we wanted to obtain participants’ uncorrected thoughts. A directed content analysis was performed using the a priori selected outcomes as themes. A deductive framework and belonging codebook had been developed prior to the interviews. Its structure and rationale were guided by information gathered during routine meetings with the study team during the control and implementation phase.³⁷ After transcription and familiarisation, the data were coded according to the codebook using the software package MAXQDA (V.10; VERBI GmbH, Berlin, Germany; table 3). Emerging subthemes were added to the respective outcomes. Qualitative data were not quantified. Field notes by the interviewer were collected before, during and after each interview, including the interviewer’s expectation of the particular interview. As these diary entries were intended for the researcher, they are not made public. In an iterative process, two researchers independently coded 12 interviews and discussed any potential differences in coding, refining the rules as necessary. After inter-rater consensus was high, the remaining 13 interviews were coded by a single researcher. Peer debriefings took place regularly with peer researchers, including discussions concerning the interpretation of data and interrelation among themes. Reporting adheres to the COREQ checklist.³⁰

Patient and public involvement

Patients and the public were not directly involved in the design of the research study. We assessed patients’ opinions of the intervention and asked specifically about the burden of the intervention and the time required to participate in the study. Our aim was to learn about the patients’ experiences and use this to inform the decision about how to incorporate the intervention into routine care. Patients were not involved in the dissemination of the results. There were several reasons for not involving patients in the process. One reason was our experience of the high burden on geriatric patients in terms of time spent on site, orientation within the clinical procedures and mental stress prior to their surgery. In addition, the focus of this research was on the early and internal implementation processes in a clinical routine that is largely dependent on the realisation of different HCP and workspaces. However, if this study had been planned and carried out in the present day, we would have included patient input throughout the study, for example, in the form of a patient project board and regular meetings and consultations with the board.

Results

Study sample

A total of 25 interviews were conducted. Nine patients were contacted. Of these, seven patients participated in the interviews. Fifteen HCPs were contacted, of whom seven were involved in the implementation process. Three HPCs were interviewed two times, once during and once after the implementation process (table 2 and online supplemental figure S1). These HPCs were directly involved in the implementation process. Due to the small sample size, no personally identifiable information is included in the quotes.

Patient interviews lasted on average 20:52 min (SD 5:30 min). HCP interviews lasted 44:18 min (SD 25:28 min). Interviews with HCPs involved in the implementation process lasted substantially longer. Quotes were translated from German to English by the authors (table 4 and online supplemental table S1). The original statements are available on request.

Table 4. Participant quotations concerning the multicomponent implementation.

No.	Participant	Theme, quotes
	Feasibility/adoption
HCP1	Nurse	“And at the end nurses simply took care of that. Then it worked without us”
HCP2	Physician	“Yes, it is definitely feasible. It only is an extremely tough process, because procedures have to be changed, that have run differently for decades and because people do no appreciate why that is”
HCP3	Physician	“… that you might [have] much better effects, if you give the patient another two weeks to optimize them. That is something, that has to be understood first, both by patients as well as colleagues and principal consultants”
	Appropriateness
HCP4	Physician	“The evidence is actually quite strong. These are mostly guideline recommendations, that simply have not yet been implemented”
HCP5	Assistant*	“I think a big factor is actually not each intervention [component] in particular, but maybe quite simple that patients feel that they are under better care”
P1	Patient	“This strength drink really gives you a boost and […] people who are really rattled - in plain language -, they would benefit maybe”
	Acceptance
HCP6	Nurse	“I had the feeling that the main obstacle was the acceptance among colleagues and that they simply did not understand why they should do it”
HCP7	Nurse	“That you slowly notice, with these never-ending [staff] education sessions, and the perpetual repetitions […] acceptance is increased”
HCP8	Nurse	“The more positive feedback came from patients, the more [HCPs] have appreciated [the procedure]”
HCP9	Physician	“And whenever something [intervention component] has been established, it is no longer a problem”
HCP10	Physician	“The nursing staff, I would say, were most accepting, simply because they are highly involved with the patients themselves”
HCP11	Physician	“Because we [anaesthetists] simply have the greatest reservations about anything that cannot be measured with a distinct laboratory or other very clear value”
P2	Patient	“Well, I am very [positive] about this thing, because I am older and maybe I will again have to depend on the anesthesia going well and that it won’t harm me”
	Sustainability
HCP12	Physician	“It has, at least in my opinion little innovational character, I would say because these are actually very normal, sensible considerations that every anesthetist should be doing for their narcosis. And in this respect, I believe it is absolutely feasible and in the ideal world should be standard”
HCP13	Physician	“Well, it is also a structural problem, and the bottom line is that time is a question of money”
HCP14	Assistant*	“There are cross-links at all. They [outpatient doctors, hospitals] all act completely independent of the other […] It is more of a structural problem”
HCP15	Physician	“And we just don’t look at the patient in the long term. And many complications, especially for geriatric patients, are not things that show up in the first 5 days. They are often complications that you see after a month or two”
HCP16	Physician	“A big problem is, I think at least here in Germany, that the health insurance companies, are sometimes quite happy when patients slip into a care level, because statutory health insurance and statutory care insurance are two different pots of money”
HCP17	Assistant*	“This is indeed a problem, because it just happens that guidelines are developed by physicians and scientists. But the reality in hospitals is largely determined by economics”
HCP18	Physician	“All medical centers calculate close at the limit”
HCP19	Nurse	“The idea of patient-centeredness is something that comes with increasing professional experience”
HCP20	Assistant*	“Go there a lot and explain. […] and that the direction comes from the very top […] that the principle says: 'here we do this like this. I want it to be done this way’. And then it gets done”
HCP21	Nurse	“In our department at times we have a turnover of about 40 people per year, some of whom come inexperienced. This means, when I give a training course in April, in December there will probably be another 30 people did not [participate in the training]”
HCP22	Nurse	“I would set up the training courses in a much more systematic way. I would put a much, much stronger focus on that [staff training]”

^*Medical assistant and medical student.

Implementation outcomes

Feasibility/adoption

Implementation in this pilot setting was mostly considered successful (HCP1). Nevertheless, the process was described as challenging, with several logistical, financial and other barriers, and with effortful persuasion required (HCP2). The short time between CGA and surgery was mentioned as a major recurring barrier. It was noted that education on the benefits of a preponed anaesthesia consultation is important for both patients and HCPs (HCP3).

Appropriateness

At the end of the study, HCPs considered the intervention to be evidence-based and appropriate (HCP4). A positive effect of the intervention was partly attributed to the side effect of greater attentiveness towards the patient (HCP5). Patients and professionals considered the intervention to be more appropriate for frailer and geriatric patients (P1).

Acceptance

Initially, incomprehension of the intervention predominated (HCP6). Over the course of the study, with more experience, many instructions and patient feedback, HCP acceptance increased (HCP7, HCP8, HCP9). Differences in acceptance were reported for ward nurses and surgery staff. Ward nurses were more accepting towards the introduction of the intervention (HCP10, HCP11). Patients generally perceived the intervention as positive (P2).

Sustainability

HCPs were confident that most of the intraoperative interventions could be routinely implemented if resources were available (HCP2). One interviewee emphasised that most of the (intraoperative) interventions are actual guidelines that already exist (HCP12). HCPs were more sceptical about preoperative interventions (HCP13). Mentioned implementation facilitators were the provision of financial resources and the simplification of the invoicing for the anaesthesia outpatient clinic. A recurring suggestion for an appropriate and sustainable implementation was to improve communication with general practitioners (GPs) (HCP14, HCP15). HCPs also mentioned the more complicated issue of insurance interests and the need for medical centres to act economically (HCP16–HCP18). Knowledge and patient-centredness were seen as the main contributors to a successful implementation (HCP19, HCP20). This was considered particularly important due to staff turnover and the high number of young professionals (HCP21). Overall, the allocation of resources, finance, time, as well as ongoing training and the need for updated standard operating procedures (SOPs) was mentioned as key factors (HCP22, HCP13).

Association between outcomes

All outcomes were highly interrelated; see for example, citations HCP2, HPC12, HCP42 (table 4 and Online supplemental figure S1 and figure 1). Estimates on sustainability were dependent on all other outcomes but did not significantly influence other outcomes. To consider an intervention as feasible depended on estimates of appropriateness and acceptance of HCPs and patients. In turn, acceptance was influenced by feasibility and adoption; for example, acceptance of outpatient prehabilitation depended on the patients’ adherence of patients (appropriateness) and sufficient room capacity for a CGA (feasibility) (figure 1).

Figure 1. Coding tree: codes and subcodes organised per barriers and facilitators. The direction of interrelation between codes is shown in the form of (dotted) arrows.

Implementation of intervention components

The order of presentation is based on the perioperative timeline, that is, components at the anaesthesia consultation first, then take-home components and finally intraoperative components (online supplemental table S1). As implementation outcomes were highly interrelated, results are reported here without strictly separating them.

Folate, cobalamin and iron substitution

Although substitution of folate and cobalamin was planned, this intervention component was not implemented. First, invoicing an additional blood sample and the substitution proved to be complicated (HCP24, HCP25), as did the assay of blood levels (HCP26). They were still administered in the trial because the protein supplementation that was eventually given to most patients contained the substances (HCP27). HCPs considered preoperative iron substitution to be a worthwhile investment, given the potential savings in banked blood and subsequent interventions in the event of complications (HCP28). In addition to the same implementation complications described for folate, another barrier was the narrow medical centre’s SOP for intravenous iron replacement (HCP29).

Medication adjustment

Adjustment of multiple and inappropriate medications was not implemented. The interval between anaesthesia consultation and surgery was very short (HCP30). Relying on the patient’s ability to know and remember their medications proved difficult (HCP31). HCPs described a lack of pharmacological expertise to identify and appropriately modify suboptimal medications (HCP32). They considered the GP better placed to address medication adjustment (and other preoperative components, eg, iron deficiency, malnutrition, frailty) (HCP33). It was hypothesised that the hospital’s medication adjustment would not be accepted by the GPs (HCP34).

Extended anaesthesia consultation

Individual risk assessment was performed in the CGA, patients were educated about delirium and pNCD and were instructed on at-home interventions (table 1). Education included general navigation of the inpatient stay, information about risk factors and the preventive nature of being physically fit, bringing familiar objects, photos, books, music and involvement of relatives. Scepticism has been expressed about the validity for the routine setting, as for the study, patients were consulted 10 days prior to surgery, which is significantly longer than in routine circumstances (HCP35). Limited room capacity and consultation time were considered problematic (HCP36). Information about the inpatient procedure was perceived positively by patients and HCPs (HCP37). Considering the extensive premedication visit and the requirement of the patient’s cooperation in the longer term before the surgery, adherence of frail patients was estimated to be poor (HCP38, HCP39). A permanent implementation into the perioperative routine was believed to solve this issue (HCP40). It was suggested to divide the premedication day into several appointments to reduce the stress on patients and room capacity. Coincidentally, it could enable the possibility for longer intervals between surgeries needed for the appropriateness and sustainability of some components (HCP41, HCP42).

Protein supplementation

The administration of protein supplements was not considered to be sustainable due to high costs and invoicing issues. Patients and HCPs rated sustainability low when asked about privately financing protein supplementation (as it is not covered by statutory health insurance) (HCP43, P3). Acceptance of and adherence to protein shakes were mixed (HCP44, P4, P5).

Physical and respiratory prehabilitation

The planned individual instruction for home-based prehabilitation proved to be beyond the scope of a feasible anaesthesia consultation. The adapted implementation included a take-home booklet with several exercises to choose from, including a diary to record which exercises were performed, when and how often before surgery. Participants were sceptical (HCP45). Adherence to home-based prehabilitation was mixed, with poorer adherence to breathing exercises (P6–P8). HCPs and some patients recognised its use particularly for more severely affected patients (P9, HCP46, HCP47). Notably, more affected patients complied less with the exercise routine (P10).

Involvement of relatives

While planning to invite patients to bring relatives or friends to the anaesthesia consultation, patients were enrolled in the study only at their scheduled visit. Therefore, it was not possible to give advance notice. Patients were then informed about the potential benefit of the presence of a confidant on postoperative outcome. Patient acceptance and understanding of the intervention was low (P11–P13).

Preoperative carbohydrate loading

Patients were informed to drink carbohydrate-containing fluids up to 2 hours before surgery to improve insulin sensitivity and reduce protein catabolism and metabolic stress response. It was highly dependent on the HCPs’ confidence regarding its appropriateness and compliance. As the common practice of promoting preoperative sobriety was internalised by patients and HCPs, HCPs initially had to be trained and persuaded (HCP48, P14, P15). Over time, they became convinced of its benefits and patients reported a positive attitude towards it (HCP49, P16).

Analgesia adjustment

Avoidance of perioperative benzodiazepine was already implemented before or during the trial (HCP50), but not the avoidance of anticholinergics, central alpha-agonists and the preferred application of short-acting muscarinic antagonists. HCPs considered it appropriate, particularly for more frail patients (HCP51).

Retainment of orientation aids

It was planned that glasses, hearing devices and dentures were to be left with the patient until anaesthesia was administered. The implementation was described as very effortful. Resistance came from the hospital’s quality management (QM), hospital hygiene, ward nurses and anaesthetists. QM was said to be concerned about potential losses and consequential liability and compensation, and the hygiene department prohibited any personal items from being sluiced into the OR (HCP52, HCP53). To address these concerns, lockers were installed in the OR. Just prior to anaesthesia, surgical nurses were to lock away the orientation aids. Surgical staff expressed concern about the additional time required for this procedure, as well as the possibility of misplacing an item during the department’s busy routine (HCP54, HCP55). It was reported that after initial significant resistance, and with positive feedback from patients, it was implemented efficiently (HCP56, HCP57). Considering appropriateness, patients and HCPs differentiated between different devices. Hearing aids were perceived as most helpful for retaining/regaining orientation, whereas extra care for dental prosthesis was perceived as unnecessary (HCP58–HCP60). While HCPs reported that patients were pleased (HCP61), patient interviewees did not perceive it as a benefit to their orientation (P16, P17). Information based on patient interviews was poor; it was only during the interviews that we realised that the patient included to evaluate the visual orientation component only needed reading glasses; the patient with the hearing aids did not want to take their aids along.

Prewarming

The use of prewarming systems before (and after) surgery has been widely adopted since an updated medical guideline emphasised its use in 2019 (HCP61). This means that prewarming has started to become routine care almost simultaneously with the start of the intervention phase of this study.

Electroencephalography-guided anaesthetic administration

The Bispectral index was planned to be given to all patients to measure anaesthesia depths and potentially reduce doses. Opinions were mixed. Some HCPs felt it was essential to use in this cohort (HCP63, HCP64). Others were sceptical about its cost-effectiveness and appropriateness in certain surgical settings (HCP65–HCP67).

Regional anaesthesia/use of epidural or peripheral catheters

Overall, HCPs viewed regional anaesthesia positively, but an informed use was recommended (HCP68). One patient commented on their negative experiences with the intraoperative interventions (P19).

Discussion

This qualitative evaluation provides a comprehensive assessment of the implementation experience of a multicomponent perioperative intervention aimed at improving postoperative outcomes and preventing functional decline in geriatric patients. After extensive preparation and HCP training sessions, the intervention was implemented during this pilot study (table 1).

Primary barriers to the implementation (process and) success were reported to be related to logistical constraints. Specifically mentioned were limited resources, short time between anaesthesia consultation and surgery, and knowledge gaps. Financing and policy constraints influenced the implementation directly (invoicing of components) and indirectly by limiting other resources, for instance, time and HCP resources, and interdisciplinary and intersectional communication. These barriers were identified in implementation studies across several settings, as summarised by Wang et al.³⁸ Flexible booking systems, adjusted distribution of health and staff resources, and easily accessible health networks have been identified as facilitators of general implementation success. The patients’ burden due to out-of-pocket payments for interventions, on the one hand, and the issue of strict delineation between inpatient and outpatient services leading to constraints on invoicing prehabilitation measures on the other, are barriers to and potential leverage points to improving perioperative care.^{39 40} The short interval between anaesthesia consultation and OR hampered the implementation of preoperative interventions; see also Hulzebos and van Meeteren.¹³ Feasibility and sustainability of the intervention were considered to be highly dependent on stakeholders’ acceptance of changes in the care process. HCPs stated that an earlier consultation relative to the scheduled surgery required a change in the mindset of both HCPs and patients. Our findings are consistent with similar studies highlighting the challenge of sustaining interventions beyond pilot settings.^{22 41 42}

As reported by Lodge et al,⁴³ some benefit of updated perioperative care was attributed to increased patient–staff interactions. The need for recurrent convincing and education of HCPs and patients for a successful implementation was mentioned by many. The gradual shift from initial HCP resistance to acceptance was perceived to be particularly evident for components that required more time (eg, CGA), for more significant procedural changes (eg, OR orientation aids) and for paradigm shifts (eg, preoperative carbohydrate loading). These findings are in line with studies describing how changes to established routines can hamper the adoption of innovations.⁴⁴,⁴⁶ HCP acceptance also increased when patients gave positive feedback. As found in similar settings, building knowledge, involving staff in the process, but also improving interprofessional communication was considered to be key factors for a successful implementation.^{47 48} In this line, staff turnover and the presence of young professionals stress the need for extensive preparatory work, ongoing training and updated and guided SOPs to ensure consistent implementation.^{19 49 50}

As described by Ali et al⁵¹ and Zietlow et al,⁵² family involvement and patient education were considered important for intervention satisfaction and positive outcomes.^{51 52} The low patient acceptance of family/confidant involvement suggests that more work is needed to engage patients in their self-efficacy. The implementation success is also dependent on patient adherence.²⁴ While patients were receptive to prehabilitation, their engagement was strongly influenced by their understanding of its benefits and their physical and social circumstances, see also Agasi-Idenburg et al.²⁵ While being the patient group, which was expected to benefit the most, frail patients were predicted to adhere poorly. The home-based prehabilitation components were met with variable patient compliance. The take-home booklet with exercise description and performance documentation as a substitute for a long on-site demonstration might increase feasibility in a busy routine care setting, but might also lead to reduced patient engagement and adherence.^{22 23}

One explanation for potentially low HCP acceptance was that the long-term effect of the interventions would not be seen by hospital staff.⁵³ The suggestion to improve communication with the GP was deemed necessary for sustainable medication adjustment, nutritional prehabilitation and reduction of blood deficiencies, but also for clinical HCP acceptance. This aligns with studies on healthcare integration, highlighting the importance of communication between care providers to improve continuity of care.^{54 55} Interesting options for future strategies include the suggestion that online resources could help to identify and safely taper polypharmacy, and that telemedical applications could be effective for prehabilitation.^{52 56} It remains to be seen if digital alternatives can improve patient-centred care in the older population.

Strengths and Limitations

Poor and inconsistent reporting of implementation strategies has been described as a problem in implementation research.⁵⁷ Here, we thoroughly described the implementation components and process to generate readily usable knowledge. To improve (the understanding of) adoption and sustainability of a wide range of perioperative interventions, we looked at the perspective of different stakeholders, including patient perspectives. Specifically, we capture a wide range of realities by including different patient groups and HCPs from various contexts and work settings. Our data represent diverse opinions and attitudes about the intervention and outcomes. Further, we focused on the routine care for elective surgery in geriatric patients in general, and our results are not limited to a specific disease or surgery procedure. It directly informs what to adapt or where to develop tailored interventions to reduce postoperative functional decline.

Our findings should be interpreted with some limitations in mind. The small sample size of each subgroup, especially the patient subgroup, makes it possible that relevant experiences may not have been captured. Also, a sampling bias cannot be ruled out, and the extent to which the results are transferable to other settings is uncertain. However, our findings are supported by the fact that statements made by participants in our sample correspond to those made by participants in other studies on the topic (Fuchs et al⁴¹ and van der Zanden et al²²). Furthermore, quantitative data on feasibility and fidelity to protocol published in Olotu et al²⁹ aligns well with our qualitative findings (eg, the implementation rate of the components aligns with the interviewees’ estimates).

Mainly for ethical and logistical reasons, we did not have the opportunity to capture the experience of the most vulnerable patients who would be expected to benefit most. Finally, no long-term data are available; thus, it is unclear how many of the components and to what extent are still found in the clinical routine.

Conclusion

This study provides in-depth contextual insights into the mechanisms of implementation in the perioperative care of geriatric patients that largely align with findings from current studies. The main difference between our and similar recent studies lies in their scope. Rather than focusing on specific components of preoperative prehabilitation, we evaluated the implementation of a comprehensive perioperative intervention, including a broad range of unique evidence-based components within the clinical routine with multiple stakeholder involvement. While many studies have focused on specific (geriatric) patient cohorts, we have included geriatric patients undergoing any elective surgery, thus generating evidence applicable to a wide range of procedures.

Future efforts should address the allocation of monetary, HCP and spatial resources, patient engagement, communication between healthcare sectors and general financial barriers.

Data availability statement

The transcriptions of the interviews are stored on a protected server of the University Medical Centre Hamburg-Eppendorf, with strongly regulated access even for study personnel. Due to substantial risk of de-identification (relatively small sample, routine care, a large amount of qualitative data, etc), individual participant data and interview transcriptions will not be shared publicly. Researchers who submit a methodologically sound proposal to the investigators that is approved by the responsible review committee will be allowed to use data. For that purpose, please contact Lisa Lebherz, I.Iebherz@uke.de.