The Role of Feedback Training on Early Postoperative Recovery and Anxiety Scores in an Ambulatory Surgical Unit: A Secular Trend Study

The present study aimed to examine the influence of anesthesia team feedback education on targeted outcomes including anxiety, pain and shivering. The research hypotheses were as follows: H1.1: Feedback education influences anxiety scores; H1.2: Feedback education influences pain scores; H1.3: Feedback education influences shivering; H1.4: QoR-10 differs between Phase 1 vs. Phase 3 of the study (see Methods section).

Included in the study were all ASA PS I-III patients, aged 18–85 years, undergoing surgery under general anesthesia in our in-hospital ambulatory surgery unit. The first 300 consecutive patients were included in Phase 1, and, another 300 patients were studied in Phase 3. Excluded were patients aged less than 18 years or older than 85 years of age; pregnant; those suffering from known cognitive or psychiatric disorders; patients suffering from opiate overuse, or addiction to recreational drugs.

The 10-item Quality of Recovery Score (QoR-10) was modified from the QoR-15 [13] and validated (see later) by our department. The maximal score for each item was 5 accounting for a total maximal score of 50 for the ten items included in the.

The ten elements of our QoR-10 score were: the patient’s ability to speak easily, the patient did not experience depression, the patient communicated easily with surrounding people, he breathed easily, he reported no mental distress, he reported no nausea/vomiting, he reported no other complaints, he moved/walked easily, he reported no shivering and reported a pain score < 4.

Pain was measured using a numerical rating scale (NRS), a generic tool by which patients describe their pain numerically, where 0=no pain and 10=worst pain conceivable. A score > 3 points triggers analgesia intervention [15].

Intraoperative pain management was tailored to the type of surgery and patient’s characteristics and included the use of fentanyl or continuous use of remifentanil, pain adjuvants such as paracetamol and dexamethasone, and when indicated and appropriate, topical anesthesia, wound infiltration or regional blocks were added. Postoperative pain assessed with NRS was managed with either morphine or meperidine, along with intravenous paracetamol or dipyrone.

Preoperative anxiety was measured using a modified (by our team – see reasoning later) Amsterdam Perioperative Anxiety and Information Scale [16]. This was performed by the attending anesthesiologist for a given patient. The anxiety from anesthesia and information about it score was not a part of the QoR-10 score, but it was recorded by the treating anesthesiologist, since preoperative anxiety could have a role in the quality of patient recovery.

The original Amsterdam Perioperative Anxiety and Information Scale (16) contains six variables, three related to anesthesia and three related to surgery:

I am worried about the anesthetic

Each variable receives from 1 (minimal anxiety) to 5 (extreme anxiety) points, accounting a total of 30 points (maximum anxiety and need for information from anesthesia and surgery). Since we could not have control over the surgeons’ preoperative education of patients, we measured only anxiety from and need for information about anesthesia, accounting a maximum anxiety from anesthesia points of 15. We considered as severe anxiety from anesthesia a score>10. This may also contain in it the patient’s need for more information about anesthesia, however for the sake of simplicity, we called it anxiety from anesthesia.

To reduce anxiety regarding anesthesia, anesthesiologists answered all patients questions regarding anesthesia in the preoperative clinic.

Does the score predict an outcome? - We measured one score to predict another one. QoR scores from Phase I were used to predict Phase III scores.

The study consisted of three phases. In Phase 1, data were collected from the records of 300 patients using the QoR-10. In Phase 2, modifiable variables in the QoR-10 were identified and targeted for change using performance feedback, during which anesthesiology team members were reminded to employ the routine clinical recommendations already applied in the department before this study, i.e. warming all patients, administering analgesics as needed and treating postoperative shivering as usual. Phase 3 included data collection from another 300 patients using the QoR-10.

In Phase 2, the anesthesiology team participated in a one-month performance feedback education program to identify ways to improve the variables that received low scores on the QoR-10. Targeted measures included treating shivering, pain, and pre-operative anxiety.

All department anesthesiologists (n=7), who were specialist-attending physicians, were instructed to adopt the following practices: warm all patients as soon as they arrive in the OR and after surgery in the PACU; pay attention to treat intraoperative pain per department routine, depending on the procedure, patient age, comorbidities, and other factors; use regional blocks and local anesthesia per anesthesiologist’s clinical judgement and routine, in addition to general anesthesia; provide detailed explanations to patients regarding the process of anesthesia, as part of the pre-anesthesia assessment, in order to reduce preoperative anxiety related to anesthesia. We did not add special drugs or techniques during this educational phase, but rather emphasized to reinforce the use of known drugs and techniques.

During Phase 3, QoR-10 scores were recorded in another 300 patients undergoing surgery in our ambulatory surgical department during a 3-month data collection period. QoR-10 scores were compared between Phase 1 and Phase 3.

In addition to the QoR-10 items and anxiety scores, the following variables were recorded for each patient in Phase 1 and Phase 3: type of surgery; type of anesthesia; comorbidities; duration of anesthesia; duration of PACU stay, complications encountered and pre-planned and unplanned admission to the hospital for overnight stay after ambulatory surgery.

It was calculated that a sample size of 296 patients in each of the data collection phases (Phase 1 and Phase 3) was required to achieve a study power of 80%; the effect size was set at 0.06, assuming that the percentage of participants with QoR-10 scores 49–50 points would increase from 77% prior to the feedback education (Phase 1) to 85% after the feedback education (Phase 3). The two-tailed alpha was set at 0.05.

All analyses were carried out using SPSS.25 software (IBM, USA). Distributions of continuous variables were tested for normality using the Kolmogorov-Smirnov test and described as mean ± standard deviation or median (interquartile range [IQR]) as appropriate to variable distribution. Nominal variables were summarized in frequency tables. Ordinal variables were tested for normality using Shapiro–Wilk test and compared by the non-parametric Mann-Whitney-U test. Continuous variables were compared between the study groups using T test for independent samples. Categorical variables were compared using Pearson Chi-square test. Potential confounders were tested in univariable analyses of associations between them and QoR-10, anxiety, shivering and pain outcomes. Confounders were controlled for in logistic analyses. All tests were two-sided and considered significant at p-value<0.05.

The final analysis was done using multivariable logistic regression.

The primary outcomes of the present study are the differences between QoR-10 scores measured prior (Phase 1) vs. after (Phase 3) performance feedback learning implemented in Phase 2. Secondary outcomes include changes from Phase 1 to Phase 3 in each of the three modifiable variables identified in Phase 1: shivering, pain, and anxiety from anesthesia scores.

All the QoR-10 scores were calculated based on the nurses’ measurements, right before the patient discharge home or (if pre-planned) to hospital ward. This timing has the advantage of further treating specific problems if this was necessary, in order to fulfill the validated Modified Post Anesthetic Discharge Scoring System criteria [17].

The main limitation of the study (but at the same time emphasizing its originality) is the lack of follow up time after the PACU state. A longer follow-up duration would eventually identify even rare adverse events and would also give an opportunity to estimate patient satisfaction with their perioperative management. Additionally, we did not find a change in the anxiety score between phase 3 and phase 1, possibly attributable to variability of explanations among anesthesiologists in the preoperative clinic, and to the lack of data regarding anxiety from surgery. This was caused by the lack of our control on surgeons’ education in phase 2 of the study.

Findings of the present study indicate that feedback education is a feasible tool for identifying actionable clinical goals, targeting them for improvement. This method may have a role in the continuing education of anesthesiologists. Additionally, the QoR-10 score can be used to assess anesthesia team performance.

However, considering the limitations depicted above, further studies are necessary to reconfirm the routine clinical usefulness of QoR-10.