Implementation of Strategies to Prevent and Treat Postoperative Delirium in the Post Anesthesia Caring Unit. A German Survey of Current Practice

Background Postoperative delirium is associated with worse outcome. The aim of this study is to understand present strategies for delirium screening and therapy in German Post Anesthesia Caring Units (PACU) in hospitals and ambulatory anesthesia facilities. Methods We designed German-wide web-based survey of 922 leading anesthesiologists in hospitals and 726 in ambulatory surgery. Results The response rate was 30% for hospital anesthesiologists. 10% (95%-confidence interval: 8–12) of the anesthesiologists applied a standardised screening for delirium. Even though not on a regular basis, in 44% (41–47) of the hospitals, a recommended and validated screening was used, the Nursing Delirium Screening Scale (NuDesc) or the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). If delirium is likely to occur, 46% (43–50) of the patients were examined using a delirium tool and 20% (17–23) screened in intensive care units. For the treatment of delirium, alpha-2-agonists (83%, 80–85) were used most frequently for vegetative symptoms, benzodiazepines in anxiety in 71% (68–74), typical neuroleptics in 77% (71–82%) of patients with psychotic symptoms and in 20% (15– 25) of patients with hypoactive delirium. 45% (39–51) of the respondents suggested no therapy for this entity. One third of the respondents indicated an age limit for pre-anesthetic medication with a mean age (SD) of 74.2 (±6.4) years and avoid benzodiazepines.

2 Abstract Background Postoperative delirium is associated with worse outcome. The aim of this study is to understand present strategies for delirium screening and therapy in German Post Anesthesia Caring Units (PACU) in hospitals and ambulatory anesthesia facilities.

Methods
We designed German-wide web-based survey of 922 leading anesthesiologists in hospitals and 726 in ambulatory surgery.

Conclusions
Monitoring of delirium was not established as a standard procedure in German PACUs. However, symptom-oriented therapy for postoperative delirium corresponded with the current guidelines.

Background
The recovery unit, also known as post anesthesia caring unit (PACU), provides ongoing nursing and medical care by specially trained personnel until the patient has completely emerged from anesthesia. In order to diagnose and treat potential adverse events, German guidelines postulate the 3 permanent presence of an anesthesiologist. By all means, a trained anesthesiologist must be on short call (1).

Delirium
Postoperative Delirium (POD) may occur, especial in the elderly, in up to 50 % of the patients and is accountable for increased mortality (2). A recent study showed that up to 19 % of the post-surgical patients developed delirium at the ward (3) and that 14 % of the patients were already tested positive for delirium in the PACU (4). Already 10 minutes after PACU admission, a pathologic RASS or NuDeSc-Score, often described as emergence delirium (ED), was associated with delirium (OR 2.4, 1.5-3.9 CI) and death after three months (OR 1.4; 0.7-3.4 CI) (5). Immediately after awakening and extubation, Monk found a 3.7% incidence of ED, declining to 1.3% when re-evaluated in the PACU (6). Another study found a 4% incidence of ED at the time of discharge (7). In contrast to ED, (post-operative) hypoactive delirium has subtle symptoms and is even more frequent. Therefore, a standardised delirium screening has been proposed by Radtke et al. (4) and the recently published European Guideline (8) also underlines screening for POD in all surgical patients. Screening should already start in the PACU and should be carried out in each shift up to postoperative day 5 with a validated delirium score (8). Besides the scientific studies that led to the development of the current guideline, there is no research or data outlining current strategies for delirium management in the PACU.
However, a detailed insight into the characteristics of clinical practice would be extremely fruitful for a successful implementation of the guidelines (9).

Pharmacological Premedication and Delirium Management
Pharmacological premedication before anesthesia and surgery reduces anxiety and improves induction conditions as well as postoperative pain management (10). Typically, benzodiazepines are administered (11). Concerning the risk-benefit assessment, the necessity of premedication for all patients has been questioned (12,13). Pharmacological premedication with benzodiazepines is associated with an increased risk of postoperative delirium (14)(15)(16), especially in the elderly.
Concerns involve the possible anticholinergic burden of premedication medication. Especially in combination with polypharmacy in the elderly, a routine premedication for these patients is open to question (17,18).
To document organizational strategies and the current practice on strategies for delirium prophylaxis, use of delirium tools for screening and therapy in German PACUs together and oral premedication with its associations subject to the clinical sector and the size of the institution, we designed a prospective, German-wide online survey.

Methods
This manuscript documents German standards in postoperative care via an online survey (19), meeting the COREQ criteria for the reporting of qualitative studies (20). anesthesiologists working in an ambulatory setting to take part in the electronic survey (LimeSurvey 2.05 software package (LimeSurvey GmbH, Hamburg, Germany). To remove responder bias, the questionnaire was re-sent once. In a short introduction we explained the significance of delirium monitoring in critical and postoperative care and the objective of the study. Informed consent was obtained before the 25 questions could be answered anonymously. The questionnaire was published before (9) and covered structural data and information about delirium screening and therapy. All participation data, necessary for an email reminder, were erased after the completion of the survey.

Statistical Analysis
In order to estimate the relevant percent or proportion, results are being presented as correlation of the answers to the total cohort together with a 95%-confidence interval, calculated by the Clopper-Pearson method. For comparing means, the Mann-Whitney-U-Test was used. Categorical data were analysed by the Χ 2 -test or Fisher's exact test, determining the level of significance as α = 5 %.

Structural Data
The

Organizational Strategies Regarding Delirium in the PACU
In 19 % (14-24 %) of the hospitals, permanent medical care was exclusively provided for the PACU.
The continuous presence of an anesthesiologists was reported in >75 % only in hospitals with more than 30 PACU beds and more than 45,000 procedures per annum.
To detect a delirious state 44 % of the anesthesiologists, who routinely score for delirium, used one of the two recommended tools at the PACU for in-as well as outpatient care: the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU; 11 % (8-16 %)) and the Nursing Delirium Screening Scale (NuDESC 12% (8-17 %). 5 % even had two tools in use (see table 3 1). That implies that two third of the anesthesiologists used drugs for premedication also in geriatric patients. Those who considered a cut-off to be reasonable used benzodiazepines rarely and preferably prescribe neuroleptics, opioids and alpha-agonists instead. Particularly promethazine was frequently used in the elderly, even though it has an extraordinary high anticholinergic burden (17,18), see table 6. The long-acting clorazepate was less frequently used in out-patient than in-patient anesthesia.

Policies for Delirium Screening in German Hospitals with a Department of Anesthesia
Asked for the preferences of all participating anesthesiologists on delirium screening in their daily practice, 77 % (71-82 %) carried out a delirium screening for at least one of the suggested groups of patients (see table 4). In contrast, one fourth would not screen for delirium under any circumstances.
Delirium screening was applied in only 46 % (40-52 %) of the patients even when delirium is suspected.

Discussion
The objective of our research was to elucidate current strategies in the handling of delirium. With 30 % for hospital anesthesiologists s, its participation rate lies within a typical range for online surveys (19) and fits response rates of the same cohort (23).
This survey shows that delirium screening rate were low in German PACUs despite the fact that today the German and European guidelines plead for postoperative delirium screening as a standard operating procedure for evidence-based practice and that several scientific studies showed a high delirium rate in the PACU (8). The results of these studies have not yet led to a cultural change in delirium management.
Only 9 % of the participating hospitals screen for delirium in the PACU. 10 % of the respondents had implemented structured programs at their facilities. 46 % used a tool to screen patients even if delirium was suspected and two thirds of the anesthesiologists did not use a delirium score in the PACU at all. In contrast, in an ICU-setting 72 % of the anesthesiologists already used a scoring system (9), which leads to the conclusion that delirium screening in the ICU is much better established as it is in PACUs. It is reasonable to expect, that with further promotion of the recently published European Guidelines an increase in screening rates will take place in the PACU as well.
The awareness of these low screening rates-elaborated in the present research-could be a first step towards improving the patients' outcome caused by delirium.

8
The second step would be to ensure the implementation of the guidelines. Therefore, information for all medical employees on delirium, e.g. with bedside teaching to the caregivers, would be necessary.
It has been documented that multicomponent programs for delirium prevention can reduce the incidence of delirium (24) and, in a geriatric population, the length of stay and hospital costs (25).
The introduction of an anesthetic protocol, designed to diminish adverse anesthetic effects, including delirium, is associated with a reduction of anesthetic recovery time (26) and thus reducing costs (about 10.80 € per minute for PACUs in a French study (27)). As a consequence, implementation of a structured protocol (28) for early delirium screening is cost-effective as it reduces PACU time and occupies fewer PACU nurses. A promising approach might be a standing order procedure, worked out by a multidisciplinary team of nurses, physicians and other experts, as the guidelines (8) suggest.
The third step would involve further raising awareness for the necessity of care for delirium among physicians as well as nurses and other caregivers. Our survey showed that in four out of five PACUs there is neither a physician permanently present nor did a structured delirium screening take place (in 1 of 10 PACUs). Anesthesiologists on duty at the PACU-continuously as in university PACUs or on short call at others-could foster these considerations by acting as role models. Therefore, interdisciplinary communication between all personnel involved in delirium care would need to be improved as it has been identified as a major issue impeding successful guideline implementation (29).

Delirium on the Ward
Only every tenth hospital provided a structured screening for delirium at the time of the survey, 1 % of the respondents practiced delirium screening in regular wards, 6 % in gerontology wards and 10 % in regular wards postoperatively.
Even when supposing that all patients at a hospital are screened, a maximum of 75 % of the patients would be tested for delirium according to a TURF analysis (21). In contrast, 25 % of the respondents would not be reached by any recommendation. Including these caregivers could be one of the biggest challenges when putting the recently published European Guidelines (8) into practice. The awareness and, if necessary, a rethinking of delirium as a sign and symptom of major complications, a strong predictor of death after artificial (30) as well as non-invasive ventilation (31), and a cause of exceedingly higher costs for the healthcare system (32) might help to launch substantiated, elaborated clinical programs for delirium management. More robust results from well-powered, randomised clinical trials should underline these findings.

Therapy of Delirium
Our study also showed that the therapy of delirium, once diagnosed, is very heterogeneous. Most of the anesthesiologists confirm a pathophysiological approach to the therapy of delirium, yet there are no strong, evidence-based recommendations for therapy of PACU delirium besides the European Society of Anaesthesiology (ESA) recommendation to titrate haloperidol 0.25 mg-wise (level of recommendation 'B'), administered by the respondents in our study to 20-80 % of delirious patients, according to their delirium subtype. Only half of the respondents treated hypoactive delirium, the subtype most frequent and at the same time hardest to identify. Consequently, one has to assume that most of the delirious patients (especially with hypoactive delirium) were not correctly diagnosed and accordingly obtained no correspondent therapy. That finding implies on what future research might put a focus.

Rethinking Routine Premedication
Not only organizational issues are essential for delirium management. According to the ESA guidelines (8), a routine premedication should be avoided in a patient older than 65 years, except for severely anxious patients (level of recommendation 'B'). Only one third of the anesthesiologists proposed a cut-off at age for premedication, which though lies nine years above the ESA recommendation.
Studies to confirm or refute the superiority of pharmaceutical benzodiazepine premedication especially in the elderly are underway (33). For a rigorous implementation of the guidelines many hospitals would have to reshape their standard operating procedures on premedication.
Limitations A qualitative online survey is a viable method for the identification of barriers to adherence to and implementation of guidelines (34). In every qualitative study, however, a bias of social desirability must be assumed, thus possibly leading to more positive findings than a neutral observation of 10 organizational practice could reveal. Since participation in an online survey can be carried out anonymously, false answers are negligible.
Neither surgeons nor nurses participated in the survey, thus, the results of this survey cannot be transferred to specialties other than anesthesiology. However, our work may overestimate the level of implementation by focusing on the expertise of leading physicians.
Our respondents represent a fifth of all leading German anesthesiologists. However, only a few colleagues working in an outpatient setting participated in the study. One needs to judge these results very carefully when drawing any conclusions concerning this group.
Our data reflect the German perspective of delirium management before the implementation of the ESA guideline in 2017. Further studies should now examine how the publication of the guideline has changed delirium management. Lastly, one would need to discuss, while taking into consideration the differences in health systems and intercultural issues, how these results might be transferable to other national contexts.

Conclusion
The aim of this study was to understand organizational practice concerning PACU medical personnel, pre-anesthetic medication, and strategies for delirium screening and therapy in German PACUs in hospitals and ambulatory anesthesia facilities. Only 10 % of the participating hospitals in our survey provided a structured delirium screening for their postoperative patients in the PACU. Where already implemented, validated scores were used. The hypoactive form of delirium was rarely treated with any medication. The current cut-off age for premedication in the elderly lies nine years above evidence-based recommendations.
Our results show furthermore that standards for delirium prophylaxis and screening in the PACU were low before the publication of the ESA guidelines (8). According to these evidence-based and consensus-based guidelines on postoperative delirium, patients should not leave the PACU without having been screened for POD and screening is recommended up to the fifth postoperative day (8).
Future studies should examine implementation rates at short intervals in order to reveal obstacles in applying these guidelines. However, every participant clearly stated her or his will to participate by activating a checkbox in the online tool.

Consent for publication
Not applicable.

Availability of data and material
The datasets used during the current study are available from the corresponding author on reasonable request.

Competing interests
BZ is the former elected president of the German Society of Anaesthesiology and Intensive Care Medicine (DGAI); STS is Associate Editor for BMC Anesthesiology. KHK, IS, VvD, TS: None.

Funding
We acknowledge support by the DFG Open Access Publication Funds of the Ruhr University of Bochum, which had neither any role in the design of the study, the collection, analysis and interpretation of data nor in writing the manuscript.      Figure 1 Cut-off for a Standardised, Medical Premedication before General Anesthesia. The figure shows at which age anesthesiologists suspend medicaments for premedication prior to anesthesia. Figure 2 22 TURF-Analysis. In a weighted TURF-analysis on delirium screening for the relevant groups recommendations for delirium screening would fit German anesthesiologists' expectations in a maximum of 221 participants (75.7%), when screening is recommended in one or more of the following settings: delirious or noticeable neurological patients, the elderly or gerontology, for in-hospital patients, particular or critical care, only postoperatively and reaches > 95% of the patients and >90% excluding "elderly" and "all patients" in a hospital cohort (220 participants, 75.3%). TURF analyses are typically used for market research.
They are useful to identify the optimal design in a combination of products to achieve a maximum of distribution (range) and sales (frequencies). Adopted on screening methods in medicine, a TURF analysis enables the examination of the total acceptance for a multicomponent measure by a combination of separate subgroups. The conjunction of the subgroups in the latter example shows, together with the graph, that even with a clear indication to establish screening for delirium even for all subgroups as described in table 4, 25% of the participants have not been reached by any recommendation -in other words, would not screen any patient. Even in the traditional fields of critical care or when delirium is suspected, 65.1% (reach) would administer a score.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. COREQ.pdf