| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Surgical, Anesthetic and Radiological Sciences, Section of Anesthesiology and Intensive Care, and Department of Health Physics, University Hospital of Ferrara, Ferrara; Departments of Medical and Surgical Specialties, and Medicine, Surgery, and Critical Care, Section of Anesthesiology and Intensive Care, University Hospital, Florence; ||Service of Anesthesia and Intensive care, Hospital Nuovo Ospedale S. Giovanni di Dio, Florence; ¶Department of Emergency, Section of Anesthesia and Intensive Care, University Hospital of Ferrara, Ferrara; #Department of Anesthesiology and Intensive Care, University Hospital of S. Chiara, Pisa, Italy.
Abstract
BACKGROUND: In this multicenter prospective study, we identified factors associated with satisfaction with anesthesia in patients staying in hospital at least 24 h after surgery.
METHODS: The study was performed in six centers. Inpatients aged more than 18 yr, who underwent a wide range of common surgical procedures, were asked to answer a 10-item instrument to measure patient satisfaction with anesthesia (mean score range, 0–10) and some specific questions, and to rate their perceived health (score, 0–10). Anesthesia staff members were invited to self-compile a Maslach Burnout Inventory.
RESULTS: The satisfaction evaluation questionnaire was returned by 1290 patients (mean age, 61 ± 16 yr; males, 54.4%). The mean global satisfaction score was 8.7 (95% CI: 8.7–8.8), being <9 in 632 (49%) and 
9 in 658 (51%) patients. The Maslach Burnout Inventory was returned by 55 anesthesiologists and 68 nurses. Multivariate regression identified five variables as significant predictors of a mean global satisfaction of 9: 1) having been treated in a service with perioperative nurses specifically dedicated only to anesthesia; 2) having been treated where anesthesia information leaflets were provided preoperatively; 3) having received more than two anesthesiologist visits after surgery; 4) having a perceived health score >8.5; and 5) being older that 70 yr. No relationship was found between staff burnout and patient satisfaction.
CONCLUSIONS: Inpatient satisfaction can be improved by an organization in which surgical suite nurses are dedicated only to anesthesia, a written anesthesia information leaflet is given during the preoperative visit and postoperative visits are enhanced.
Patient satisfaction with anesthesia, i.e., the balance between expectations and perception of what was received, is of concern (1), and staff must continue to identify, monitor, and modify factors that may improve it.
To identify the many factors involved requires a large sample (2), but one from a single center could create bias and take a long time to run, during which both anesthesia and surgical practices might change. A multicenter study is preferable, provided that each center recruits enough patients (2), although overall greater variability may arise.
Considering anesthesia as the exposure and patient satisfaction as the outcome, investigators focus on the exposure–outcome relationship, but the exposure (anesthesia) adopted for each patient depends on many factors, some of which are interactive; any variable associated with both exposure and patient outcome in a dataset is a potential confounder (3). Such potential confounding variables can be classified as follows: i) patient-related: gender and age (4–6), education (7), anxiety and depression (8), perceived health (7), and comorbidities (4,5); ii) surgery-related: type, extent, and duration of surgery (4); iii) center-related: process management of the anesthesia service and hospital concerning, e.g., preoperative visit (9), and postoperative analgesia (10), together with human factors involving the staff (11) and especially the nurses (12,13). As a result, the conceptual framework of the interconnection between variables potentially associated with patient satisfaction and anesthesia in multicenter studies is unique.
The aim of this multicenter prospective study was to identify factors associated with satisfaction with anesthesia in a population of patients staying in hospital for at least 24 h after surgery.
METHODS
The study was performed on data collected by six anesthesiologists responsible for the management of six different surgical suites (Centers A through F) in four hospitals (three teaching and one nonteaching), located in three North-Central Italian cities. The number of beds per hospital ranged from 330 to 1700, and the annual number of anesthetic procedures from 4700 to 19,000, each center having from two to eight operating suites. The number of anesthesia specialists working on a regular basis (excluding night shifts) ranged from 6 to 25.
From our experience with the same instrument (6), we deemed that 1130 patients treated within approximately 3 mo were required to detect a 20% difference between any two groups of patients differing in one variable (
= 0.05 and power = 0.80).
All adult patients aged >18 yr, who consecutively underwent elective surgery between September 20 and December 20, 2004 and remained in the hospital for least 24 h postoperatively, received a questionnaire. Patients who declined to participate, who did not speak Italian, were confused at the time of the postoperative visit, or for whom intensive care unit admission was planned after surgery, did not receive the questionnaire.
The study was approved by the Ethics Committee of the coordinating center and each participant gave informed consent.
Patient Questionnaire
Each patient received an anonymous form in three parts approximately 1 day after surgery. The first part was completed by the investigator, who was not a member of the anesthesia team taking care of him/her. The patient was then required to complete the second and third parts and give it to the head of the ward nurses before leaving the hospital. As a rule, the investigator had to offer assistance to the patients who agreed to participate, reading the questions and, when required, recording the answers on the form.
In the first part of the patient form, the investigator recorded the following variables: gender, age, date and duration of surgery, ASA classification, type and extent of surgery (minor, moderate, or major (14)), type of anesthesia (regional, general, or regional plus general), and postoperative analgesia management (managed by the anesthesiologist, or according to ward rules, or patient-controlled).
In the second part of the form, each patient had to answer a 10-item instrument, validated in the same language and country, to measure satisfaction with recent anesthesia (6). Items were grouped in the following domains: physical ("Pain at the site of surgery," "Vomiting and Nausea"), emotional ("Feeling of well-being," "Feeling safe," "Feeling relaxed," "Feeling anxious or frightened"), and relational ("Information given by anesthesiologist," "Attention to the patient," "Kindness/Regard of caregivers," "Demands promptly answered"). Using a Numerical Rating Scale with the words "no satisfaction" [0] and "the maximum satisfaction possible," [10] respectively, each patient was asked to indicate the level of satisfaction with each of the 10 items. Mean satisfaction scores ranging from 0 to 10 were subsequently derived for each domain and for all items (global): the higher the score, the higher the satisfaction. Subjects were also asked to record the level of their perceived health on a similar 0–10 Numerical Rating Scale.
In the third part of the form, information from the patients was collected, which included: date, number of years of education (categorized as <8, 8–13, and >13), suffering anxiety or depression (no/yes), explanation received before recent surgery (no/yes), and number of preoperative and postoperative visits performed by the anesthesia team associated with the recent surgery (both categorized as none, 1, 2, 3, or more).
Anesthesiologist Questionnaires
In each center, the anesthesiologist responsible for the study completed a questionnaire on the hospital, and the anesthesia service (whether there were perioperative nurses specifically dedicated only to anesthesia, whether the service included pain care, whether a patient anesthesia and/or analgesia information leaflet was supplied during the preoperative visit). Perioperative nurses were considered to be those specifically dedicated only to anesthesia when taking care of patient preparation and anesthesia administration regularly, and not acting as scrub or circulating nurse on the basis of the daily staff needs. An anesthesia information leaflet was considered a written instrument to improve patient knowledge about types and phases of anesthesia. The characteristics of the leaflet, which were used in Centers B and F, are reported in Table 1; the main differences between the two leaflets were on the typographical aspect and the type of approach (more friendly in Center F). An analgesia information leaflet was defined as preparing patients for postoperative pain measurement and treatments.
|
Each anesthesiologist and nurse who worked regularly in the surgical suite received a Maslach Burnout Inventory (MBI) form (15). This MBI questionnaire consists of 22 items allowing the three components of burnout (emotional exhaustion, depersonalization and personal accomplishment) to be computed, and required individuals to choose one of the following answers to each statement: never (score 0), a few times a year (score 1), once a month (score 2), a few times a month (score 3), once a week (score 4), a few times a week (score 5), every day (score 6). There were recommended cut-off scores, with burnout being present when scores were high for emotional exhaustion (7) and depersonalization (10) and/or low for personal accomplishment (
33), according to the validation studies performed in the country (16,17). These answers were anonymous. Questionnaires were collected in a closed box for the first 2 wk; subsequently, the anesthesiologist sent the box to the central study coordinator (M.C.), who entered the answers with the respective center codes in an "ad hoc" computer database.
Data Analysis
Data collected from the patient forms were entered by the individual centers and sent to the central coordinator, who matched the answers for each patient with data on the center involved. The incidence of burnout reported by the staff of each center, arranged in rank order, was then related to the patients in that center.
Statistical Analysis
Continuous variables are reported as mean ± 1 standard deviation. The scores recorded by Numerical Rating Scale (satisfaction with anesthesia and perceived health) are means with 95% confidence interval (95% CI).
Statistical analysis was performed using the software packages SPSS v. 11.5 and Statgraphics v.4.0. A Type I error <5% in two-tailed tests was considered significant.
A 
2 test, or Fisher’s exact test when appropriate, was used for categorical comparisons. Owing to the distribution characteristics, comparisons among the MBI scores reported by the staff of the centers were performed by Kruskal-Wallis, or ANOVA and Tukey post hoc test.
Global satisfaction (10-item) and the three domains did not show a normal distribution according to the Kolmogorov-Smirnov test (P < 0.000), making nonparametric analyses suitable (Mann-Whitney or Kruskal-Wallis). Moreover, multiple linear regression, although attractive, was not suitable because most of the independent variables were categorical and the satisfaction scores were mainly in the high value extremity. Instead, multiple logistic regression models (backward, forward stepwise, and enter methods) were used with patient satisfaction data as the dependent variable. To make it categorical, the median satisfaction score, i.e., 9.0, was used as the cut-off value. Univariate regressions were first used to identify variables significantly associated with higher patient satisfaction, which were subsequently introduced into multivariate models to evaluate their contribution. The deviance reduction given by the model was evaluated by the MacFadden index (range, 0–1), which is directly affected by the number of variables.
To simultaneously consider both patient- and center-related factors, all variables were entered together for final logistic analyses.
RESULTS
Patient Questionnaire
One-thousand-five-hundred-six patients (range per center, 151–362) met the inclusion criteria, of whom 58 refused to participate, and 30 returned blank forms, giving a sample of 1290 patients. The mean percentage of completed forms returned was 89.1%, range 77.7 (Center F) to 93.3 (Center D).
The number of forms lacking at least one of the items required to compute a satisfaction score was 28, and 19 (68%) of them missed only 1 item. No form was returned with >50% of the items for satisfaction omitted. The third part of the patient form was not collected in Center F, and nine patients from Center E did not complete it. The median time elapsed between anesthesia and questionnaire answer was 1 day in all centers except Center E, where it was 2 days.
The 1290 patients studied had undergone a wide range of common surgical procedures: orthopedic (35.6%), urological (29.8%), abdominal (12.2%), endocrine (7.8%), vascular (6.8%), gynecological (3.8%), thoracic (2.2%), and other (1.9%). The mean age was 61 ± 16 yr (median, 65; quartiles, 51–73) and 702 of the patients (54.4%) were males.
Anesthesiologist Questionnaires
The percentage return of MBI questionnaires from centers was 60%–100% for anesthesiologists and 65%–100% for nurses. The numbers of items not completed were one in eight questionnaires, two in one, and three in three questionnaires. Table 2 shows scores for emotional exhaustion, depersonalization and personal accomplishment, and the incidence of burnout. Emotional exhaustion was significantly lower in Center F only in comparison with Center E in the post hoc test. The general incidence of burnout was lowest in Center C but was not statistically significant when considering the number of centers.
|
Data Analysis
The mean global satisfaction score was 8.7 (95% CI: 8.7–8.8), being 8.7 (95% CI: 8.6–8.7) for physical, 8.6 (95% CI: 8.6–8.7) for emotional and 8.9 (95% CI: 8.8–8.9) for relational domain. The scores of physical, emotional and relational domains, and global (10 items) satisfaction reported according to patient characteristics, anesthesia, and center practices are given in Table 3. No gender difference was found. All domains, as well as the global score, were significantly different when the patients were grouped according to the other variables. The Cronbach’s was 0.9119 for all 10 items and 0.8403 for the three domains. When the sample was split according to the median global value, 632 patients (49%) reported a value <9 and 658 patients (51%) 9.
|
The variables concerning patient and surgery in relation to the lower or higher satisfaction recorded (mean global satisfaction score <9 or 9, respectively) are reported in Table 4. Table 5 shows the variables concerning patient anesthesia and center practices associated with lower and higher satisfaction by univariate analysis.
|
|
Multivariate models revealed five variables as the main significant predictors of higher satisfaction with anesthesia (Table 6). These were: having been treated in a service with nurses dedicated only to anesthesia; having been treated in a service that provided an anesthesia information leaflet during the preoperative visit; having received more than two anesthesiologist visits after surgery; having a perceived health score >8.5; and being older than 70 yr. The Mac Fadden index for the global score was 0.17, not too different from those of the three domains: 0.12 for physical, and 0.15 for both emotional and relational domains.
|
When all variables, patient- and center-related, entered final regression analyses, none was found to be significantly associated with higher patient satisfaction in more than one domain, except those previously identified (reported in Table 6), and education, which was significant on physical and emotional domains, and only for the patients reporting a number of years >13. Nevertheless, education was not significant at the univariate analyses (Tables 3 and 4).
A multivariate logistic analysis performed on data collected in five centers, to exclude a center effect due to the presence of two of the main significant predictors in Center F (i.e., having nurses dedicated only to anesthesia and providing anesthesia information leaflet), confirmed the relationship with the main significant predictors identified in the global sample.
DISCUSSION
This is the first multicenter study, which shows that having nurses dedicated only to anesthesia, giving an anesthesia information leaflet during the preoperative visit, and visiting the patient more than twice after surgery are independently associated with higher patient satisfaction with anesthesia. Nurse shortage and funding restrictions have compelled some hospital managers to make the work of perioperative nurses more flexible. As a result, each nurse takes care of anesthesia or acts as a scrub or circulating nurse according to daily needs. Our study demonstrates the positive effect on patient satisfaction of nurses dedicated only to anesthesia, as suggested by others (18).
Video and/or printed information decreases the anxiety level before anesthesia (19), and are used by more than half of United Kingdom anesthetic departments (20). Nevertheless, detailed anesthesia drug information is excessive (21), and an information campaign did not modify the problems reported by patients in three Swiss hospitals (22). In the last study, one hospital added information to a leaflet already present, one expanded the preanesthetic care unit and handed out the leaflet in preoperative settings, and the third one introduced the leaflet used in the first hospital (22). Only in the second hospital there was a significant reduction of the percentage of patients who reported that a problem was present in the domain of information. Nevertheless, dichotomizing answers (problem present or not) could have reduced the ability to detect any small effect of those interventions. In our study, the comparison was performed between centers distributing the leaflet or not, and the characteristics of the leaflets used were similar.
Receiving more than two anesthesiologist visits after surgery significantly affected our patients’ satisfaction, despite the small number [76]. Zvara et al. (23) did not find this effect in patients who received one, two or three postoperative visits. The difference might be attributed to the ceiling effect of the satisfaction scale used in (23) and/or to the exaggerated ratings reported after one visit.
Logistic regression analysis also identified two nonmodifiable patient-related variables that influenced patient satisfaction: perceived health and age, on which there were reports ((7) and (4–6), respectively). An effect of education level, already noticed (7), cannot be completely excluded, although lower in significance than that of the previous factors. The lack of relationship between staff burnout and patient satisfaction also reported in Canadian (12) and United States hospitals (13) could be due to the shorter time spent with conscious patients by the anesthesia staff. However, the MBI subscale values were similar to those reported by anesthesiologist from other countries (24).
Some variables were significant in the univariate analysis but not in the multivariate model. The lack of association between type of anesthesia and patient satisfaction has been reported recently by others (25). Regarding postoperative analgesia, the low number of patients receiving patient-controlled analgesia and the possible administration of rescue analgesics, not recorded, in some patients could have influenced the results. The explanations reported by the patients as received in the operative room may have been surrogates for other kind of information. Some patients may have recorded explanations without remembering exactly whether they received information verbally and/or as an information leaflet, in the operating room or at the time of an anesthesiologist’s visit before surgery. Finally, the lack of effect of being treated in a service including pain care could have been due to the "pain relief" campaign of the Italian Health Ministry.
With respect to the strengths and limitations of this study, it was conducted on a voluntary basis by a research group moved by a cohesive spirit, to achieve a shared research goal, without any financial remuneration. The two Italian regions where the study was performed have the same National Health Service, funded by taxpayers and are similar in health care and population. A wide range of inpatient surgical procedures was studied, as in a similar investigation (5). The instrument used to assess patient satisfaction with anesthesia was developed according to the item relevance recorded by 100 patients, and had previously undergone formal validation (6). Moreover, internal consistency (Cronbach’s ) demonstrated the reliability of the instrument.
Among the limitations, we observed a possible bias in ratings owing to the assistance given to some patients in completing the questionnaire (1). We did not record the number, but this assistance can increase the response rate, reduce the number of missing items, and influence answers due to patient social desirability. To minimize that bias, the investigators were not involved in patient care. Another potential source of bias was patients declining to receive the form or returning it blank, a common problem in all questionnaire studies. Moreover, outpatient and specialized surgery, e.g., cardiac, oral and intracranial, were not included; the range of surgical procedures considered excluded some specialized operative settings. Finally, our aim was to identify factors predictive of high satisfaction levels, and nothing can be concluded about dissatisfaction.
Our study demonstrates that three easily modifiable variables are significant predictors of high satisfaction with anesthesia, suggesting that inpatient satisfaction can be improved by an organization in which surgical suite nurses are dedicated only to anesthesia, an information leaflet about types and phases of anesthesia is given during the preoperative visit and postoperative visits are enhanced.
ACKNOWLEDGMENTS
The authors are indebted to Maddalena Pasini and Tamara Quaranta for their invaluable help with data collection, and to Alessandra Bassani for helping with form management.
Footnotes
Accepted for publication May 1, 2007.
Supported, in part, by a grant from the Ministero Italiano dell’Università e della Ricerca (MIUR).
Address for correspondence and reprint requests to Dr Maurizia Capuzzo, Dipartimento di Scienze Chirurgiche, Anestesiologiche e Radiologiche Sezione di Anestesia e Rianimazione Azienda Ospedaliera S. Anna Corso Giovecca 203, 44100 Ferrara, Italy. Address e-mail to cpm{at}unife.it.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
