Secondary Logo

Journal Logo

Original Articles – General

Evaluation of physical and mental recovery status after elective liver resection

Arnberger, Michaela,b; Vogt, Andreasa; Studer, Peterc; Inderbitzin, Danielc; Pulver, Caroled; Röhrig, Bernde; Jakob, Stephan Md; Greif, Roberta

Author Information
European Journal of Anaesthesiology: July 2009 - Volume 26 - Issue 7 - p 559-565
doi: 10.1097/EJA.0b013e328328f552
  • Free



Quality of recovery was recommended as an important endpoint of outcome after anaesthesia and major abdominal surgery [1–3], besides morbidity and mortality. In addition, short-term and long-term qualities of recovery might be influenced by complications that are probably a result of multiple variables related to patient, surgery and anaesthesia. The preoperative prediction of postoperative complications is therefore difficult [4]. A number of studies have attempted to define variables to improve outcome and to reduce costs that occur due to complications [5–7].

For the evaluation of the recovery process, standardized and valid instruments should be used [8,9]. Also, definitions of outcomes vary among studies, which limits useful comparisons between them [10,11]. The Short Form 36 Health Survey (SF-36) is a validated instrument for testing quality of life [12,13]. It determines eight subscales of quality of life: physical and mental component summary scores [14], which might be important parameters for recovery. In one clinical trial, the SF-36 was used to evaluate epidural analgesia after colonic surgery [15]. In contrast, the Quality of Recovery 9 (QoR-9), a nine-item questionnaire, was designed for the assessment of postoperative recovery with a validation in day case and inpatient surgical settings [3,16].

In summary, SF-36 and QoR-9 are useful patient-assessment tools for clinical trials during daily clinical care [8,16–19], but these outcome measurements have rarely been used in the setting of anaesthesia for major abdominal surgery [20,21] and data concerning operating expense are lacking.

With this prospective, clinical observational trial, we evaluated the feasibility of the SF-36 and QoR-9 for the assessment of 1 week of recovery after liver resection surgery and prediction of complications.


After regional ethics committee approval (no. 017/06) and written informed consent, we included 19 patients with ASA physical status I–III and age 18–80 years, scheduled for a partial liver resection in our prospective, observational clinical pilot trial, from April 2006 to June 2007. Patients who were not native German speakers or those with any preoperative cognitive dysfunction were excluded.

Short Form 36 Health Survey and Quality of Recovery 9

SF-36 is a generic measurement of perceived health status that incorporates behavioural functioning, subjective well being and perceptions of health [22]. It is a multipurpose short-form health survey with 36 questions. It yields an eight-scale profile of functional health and well being scores: limitations in physical activities because of health problems; limitations in role activities due to physical health problems; bodily pain; general mental health perception; vitality in combination with energy and fatigue; limitations in social activities; limitations in usual role activities because of emotional problems; and general mental health [22,23]. Labels and items of the SF-36 are summarized in Table 1.

Table 1:
Short Form 36 Health Survey: 36 labels with content, eight subscales and two component summaries

The test also includes two summary component scores for physical health [physical component summary (PCS)] and mental health [mental component summary (MCS)]. Each summary score is derived from four dimensions: the PCS is derived from physical functioning, role activities due to physical health problems, bodily pain and general health. The MCS is derived from the remaining four: vitality, social functions, role activities due to emotional problems and general mental health. The SF-36 for self-administering and 1-week recall was recorded the day before surgery as baseline and at postoperative day (POD) 7.

All 36 items of the completed SF-36 questionnaires at baseline and POD7 were recorded and then processed on a personal computer with the Hogrefe TestSystem software (Hogrefe TestSystem 3.8.4., Hogrefe Verlag, Goettingen, Germany, 1998). This software is programmed to calculate according to a complex algorithm the eight subscales based on the results of the 36 items. In addition, there is an algorithm to determine PCS and MCS from the corresponding subscales. These calculations could be done manually, but this has the drawback of being very time consuming even for an individual patient. For the output of the results, the Hogrefe TestSystem software offers a z-transformation of the data; for example, the results of an individual patient are referenced to a sex-matched and age-matched, German normal population. The database of this population is integrated in the Hogrefe TestSystem software. Hereby, a z-value of zero means no deviation from this normal population; a negative z-value means a lower quality of life; and a positive z-value means a higher quality of life.

The QoR-9 questionnaire consists of nine items, each scoring 0, 1 and 2 points, with a maximum of 18 points (0 is the worst and 18 is the best count) [17,24]. The nine items of the QoR-9 and the scoring system are shown in Table 2. The QoR-9 was recorded at baseline and at POD1, POD3, POD5 and POD7 always in the evening, so that the same time intervals were used. In the German version, the first six questions are exactly the same as in the English version, but the last three questions are adapted to the German language for better understanding; therefore, the rationing scheme of these questions is inverted, keeping the same content as in the English version [24].

Table 2:
Quality of recovery questionnaire (Quality of Recovery 9): nine items with answers

For both tests, we measured the duration of the interview in minutes and the difficulty in answering the questions on a continuous visual analogue scale (VAS 0–100 mm; 0 for no difficulty and 100 for highest difficulty) [25]. SF-36 and QoR-9 tests were filled in by the patient (self-administering). The investigator remained in the patient's room during the test. If the patient was too weak to read the text, the investigator helped only by reading the questions.

Anaesthetic procedure

All patients received standardized combined epidural–general anaesthesia between the seventh and 10th thoracic interspace using bupivacaine 0.25%, 0.1 ml kg−1 h−1 during surgery. Anaesthesia was induced with propofol, fentanyl and rocuronium and maintained with isoflurane in 35% oxygen. Postoperative pain was treated by an infusion of bupivacaine 0.1% and fentanyl 2 μg ml−1 plus norepinephrine 2 μg ml−1 at an amount of 2–15 ml h−1 epidurally.

Surgical procedure

The surgeons followed the general principles of liver resection surgery [26]. All patients were operated on or supervised by senior surgeons, according to departmental guidelines. All patients were admitted to the ICU after surgery.


Demographic and morphometric data, additional perioperative data (details of surgery and anaesthesia) and duration of total hospital stay (LOS) were recorded for each patient. Sepsis-related Organ Failure Assessment (SOFA) score, Simplified Acute Physiology Score II (SAPS II) and risk of hospital death were recorded after ICU admission.


The complications were recorded by an independent observer in the visceral surgery department from the patient documentation system during patients' hospital stay. A complication was defined as any event that required treatment measures not routinely employed in the postoperative setting of patients after liver resection.

The complications were classified into eight groups: cardiac, respiratory, thrombotic (central vein and peripheral vein thrombosis), infectious, gastrointestinal, neurological, nephrological and surgical, such as ascites and secondary bleeding [27,28].

Baseline values for the eight dimensions of SF-36, MCS, PCS and QoR-9 were compared between the groups with and without postoperative complications. The existence of a complication was recorded as either yes or no; if a patient had more than one complication it was counted as one.

Statistical analysis

Data collection, data management and statistical analysis were performed with statistical package SPSS version 15. For metrical variables, means and SD or quartiles or both were calculated. Data are presented as means ± SD, quartile 1 and quartile 3 and 95% confidence intervals (95% CI). The Wilcoxon signed rank test was used to compare the two questionnaires (SF-36 and QoR-9) and the time-dependent differences within the questionnaires (e.g. from baseline to POD7). For group comparisons of independent groups (e.g. patients with and without complications), the Wilcoxon rank sum test was used. The Wilcoxon rank sum test was also used for the comparison of LOS in the groups with and without complications. For comparison of the eight dimensions of the SF-36 test, PCS, MCS and QoR-9 at baseline between the groups with and without complications (all parametric data), the Student t-test was used. For assessment of the dependence of PCS, MCS and quality of recovery compared with an increasing ASA class, the Jonckheere–Terpstra trend test was used. The analyses of data were done in an explorative manner and the outcome of a statistical test with a P value less than 0.05 was considered significant.

Effect sizes (d) were calculated for the eight dimensions of the SF-36, PCS and MCS of the SF-36 as follows: [mean POD7 − mean baseline]/pooled SD. For the QoR-9, summary scores of d were calculated as follows: [mean POD1 or POD3 or POD5 or POD7 − mean baseline])/pooled SD. In addition to d for the eight dimensions of the SF-36, PCS, MCS and QoR-9, summary scores at baseline for the groups with and without complications were determined according to the formula: [mean group with complications − mean group without complications]/pooled SD.

Effect size was classified in a positive or negative direction as no effect for less than 0.2; as low from 0.2 to less than 0.5; as moderate from less than or equal to 0.5 to less than 0.8; and as large for more than 0.8 [29,30]. For graphic illustration of the distributions, box-whisker plots were used.


The demographic baseline characteristics and preoperative clinical and intraoperative data are summarized in Table 3. Epidural catheters were inserted in 18 patients, one had patient-controlled intravenous analgesia postoperatively because of a failure to place the epidural catheter.

Table 3:
Baseline characteristics and intraoperative data of patients

Five patients had a benign liver disease and 14 had a malignant liver tumour. In all 19 patients surgical incision was at the right costal arch. Right hemihepatectomy was performed on 11 patients, left hemihepatectomy on three, central hemihepatectomy on one and segment resections on four patients. Out of 19 patients, only three needed intraoperative blood transfusions: two needed two units and one six units.

All 19 patients filled in the two questionnaires preoperatively. On POD7, 15 patients answered both tests. Two patients were reintubated in the meantime and later died after multiorgan failure, and two case report forms of the SF-36 and QoR-9 were lost during the ICU stay. At baseline and 1 week after surgery, the QoR-9 questionnaire was shorter and easier to fill out than the SF-36 (Table 4). Between baseline and POD7, difficulty in answering the QoR-9 items increased significantly.

Table 4:
Feasibility data for Quality of Recovery 9 and Short Form 36 Health Survey

Short Form 36 Health Survey

Out of all the questionnaires, 89% were answered (n = 34), 11% were not available (n = 4, re-intubation and later death, two lost). The time course of the eight SF-36 dimensions is shown in Table 5, and the individual differences between POD7 and baseline are presented in Fig. 1. PCS but not MCS scored significantly lower at POD7.

Table 5:
Time course of Short Form 36 Health Survey z-data in physical and mental health components
Fig. 1

PCS and MCS decreased with increasing ASA class. At baseline, the PCS in patients with ASA class I was 0.01 ± 1.05, with ASA class II it was −0.35 ± 1.34 and with ASA class III it was −0.84 ± 1.56 (P = 0.341). At baseline, the MCS in patients with ASA class I was 0.24 ± 1.17, with ASA class II it was −0.69 ± 1.64 and with ASA class III it was −1.69 ± 1.03 (P = 0.062).

Quality of Recovery 9

Of all QoR-9 questionnaires (n = 80), 84% were answered completely and the other 16% (n = 15) were not available. The time course of the QoR-9 summary score is shown in Table 6 and the averaged individual differences between POD1, POD3, POD5, POD7 and baseline are presented in Fig. 2.

Table 6:
Time course of Quality of Recovery 9 summary scores
Fig. 2

Quality of recovery in patients was lower in ASA class III than in ASA class I and II. At baseline, the quality of recovery in patients with ASA class I was 16.6 ± 2.2, with ASA class II it was 16.7 ± 1.5 and with ASA class III it was 14.4 ± 2.0 (P = 0.055).

Postoperative complication

Demographic and morphometric data were similar for patients with and without complications at baseline. Of all patients, 36% had no complications (n = 7). A total of 33 postoperative complications (one cardiac, three respiratory, four thrombotic, seven infectious, four gastrointestinal, three neurological, four nephrological and seven surgical) were found in the other 12 patients during their hospital stay [31]. The LOS was 25 ± 13 days for patients with complications and 12 ± 2 days for those without complications (P = 0.002).

The baseline PCS z-value was significantly lower with a high effect size in patients with complications (n = 12) than in patients without complications (n = 7) (−0.76 ± 1.46 vs. 0.27 ± 0.56; P = 0.044, d = −0.84). No difference was found for the baseline MCS z-value (−0.61 ± 1.37 vs. −0.76 ± 1.63, P = 0.831, low d = −0.10) and for the baseline QoR-9 summary score (16.57 ± 1.81 vs. 15.7 ± 2.09, P = 0.384, low d = −0.44). Not a single one of the eight baseline dimensions of the SF-36 test correlated significantly with complications, nor was there a correlation with the SOFA score (P = 0.340), SAPS II score (P = 0.522) and risk of hospital death (P = 0.483). The weight of the resected liver was 719 ± 499 g in patients with complications compared with 738 ± 223 g in those without complications (P = 0.931).


In this prospective, clinical pilot trial, the SF-36 suggests that baseline physical and mental health of patients before elective liver resection might be lower than that in the general population and that physical health is affected more than mental health 1 week after surgery. QoR-9 determines feasibly the time course of recovery with a return to baseline within 7 days. The group with complications scored significantly lower for the baseline PCS than the complication-free group and thus preoperative impaired physical health might be predictive for postoperative complications.

The SF-36 is a feasible and reliable instrument with sufficient discriminatory power to detect changes in the quality of life recovery of patients in the postoperative setting [15]. Our results for the SF-36 suggest a significant change in the physical but not in mental behaviour of the patients after liver resection based on the summary scores. In this kind of patient, mental recovery seems to be easier than physical recovery. That is surprising because lots of patients with end-stage liver disease have mental deterioration due to liver disease.

The QoR-9 scores returned to preoperative values at approximately the seventh postoperative day after a significant decrease from POD1 to POD5 in this inpatient population after liver resection. The QoR-9 was the most cited instrument to assess recovery [3], but it was mostly used for ambulatory surgery. It was recommended for postoperative recovery and audit for quality assessment purposes [3]. The time course in our inpatient population confirms the earlier reported responsiveness to change of the QoR-9 in the ambulatory setting. As the quality of recovery in this pilot trial comprehensively measures the influence of patients' overall health status on well being, it may also provide a summary measure of current and past disease severity.

The shorter QoR-9 test works better during the first week of recovery and detects differences during that period easily because of a short recall time, which allows repeated measurement in this period. In addition, it was easier and took less time for patients to answer these questions compared with the SF-36. The SF-36 with the shortest recall time of 7 days is not designed to detect fast changes within 1 week. But both tests were feasible to assess quality of recovery after major abdominal surgery, without overloading patients who had recently undergone major surgery and also giving the staff an easy to handle instrument during daily clinical care in the preoperative and postoperative setting [8,32].

The time to fill in a questionnaire is one aspect of its acceptance: the longer it takes, the less acceptable it is [9]. Weinberger et al. [32] assessed the time required to fill in the SF-36 by two different methods of administration. Self-administration required about 13 min, whereas a trained interviewer took only 10 min. The shorter QoR-9 takes 2–3 min to fill in by self-administration [17]. We found fill-in times were comparable to those of others in a major abdominal surgical patient population [32].

Quite interestingly, we found even in this observational trial with a limited sample size that the baseline PCS reflecting physical health was significantly lower with a high effect size in the group with complications compared with the group with no complications. No difference could be found for baseline MCS or the QoR-9 score taken prior to surgery. However, Anthony et al. [4] found preoperative differences between patients with and without complications after colon surgery for mental and social functioning.

Interestingly, we found a trend for lower QoR-9 scores (P = 0.055), MCS values (P = 0.062) and PCS values (P = 0.341) at baseline with increasing ASA classes. For hepatic resection surgery, the only independent predictors of perioperative morbidity and mortality were blood loss and the number of hepatic segments resected [33]. However, ASA classes were not reported in that study population. Blood loss and number of hepatic segments resected are intraoperative parameters. Although the ASA score, predictive of anaesthetic complications, can be determined preoperatively, it is a single physical status score [4]. The SF-36 and quality of recovery scores are based on several different factors covering more dimensions of well being than single physical status scores and also have the advantage of preoperative determination. In contrast, no difference between the complication and no-complication groups was found concerning the eight scales of the SF-36, the SOFA score, the SAPS II score and the risk of hospital death. Our prospective, observational study, however, was not powered to detect such correlations and further larger, prospective clinical trials are needed to evaluate this interesting relationship between physical and mental health and QoR-9 score with these parameters.

Because the P value provides only an indication of the presence or absence but not the degree of the observed effect [30], we calculated effect sizes to evaluate both magnitude and direction of effects, for example surgery and anaesthesia or complications. It is currently the most accepted variable for determining the importance of a group or an individual change in health-related quality of life measurements and is independent of the number within one group and therefore appropriate to small study groups [30,32,34]. To determine the effect of surgery and anaesthesia, our study corresponds to a single group before (baseline) and after (POD7) study design [34].

A limitation is the lack of power with the inclusion of only 19 patients and the missing values due to loss of follow-up or medical reasons such as intubation. However, 50% of the missing data were randomly lost and data analysis per protocol (‘best cases’, analysis of patients without missing data) yielded no relevant difference to the all-cases analysis. The effect sizes (independent of sample size) for the repeated measurements and the group comparison (complications vs. no complications) were mainly in the large range. Thus, valuable hypotheses for further larger clinical studies can be generated from these data.

In conclusion, the SF-36 and QoR-9 are feasible and valuable tools to evaluate perioperative physical and mental status and recovery after elective liver resection. The SF-36 suggests that major abdominal surgery such as hepatic resections has a higher impact on physical health than on mental health after 1 week, whereas short-term recovery with a baseline return of 7 days can be feasibly determined with the QoR-9 due to the short recall time. Preoperative impaired physical health might be predictive for postoperative complications, which have to be proven in a future, larger outcome study.


We would like to thank the entire anaesthesia and ICU staff of the University Hospital Bern who supported the conduct of the study. Furthermore, we would like to thank Volker Hartwich, C.R.N.A. in the Anaesthesia Department, and Judith Kaufmann, R.N. in the Department of Intensive Care Medicine, without whose tremendous effort none of the provided data could be collected and analysed. Special thanks also to Jeff Crowder for the revision of the manuscript in English.


1 Gill TM, Feinstein AR. A critical appraisal of the quality of quality-of-life measurements. JAMA 1994; 272:619–626.
2 Guyatt GH, Osoba D, Wu AW, et al. Methods to explain the clinical significance of health status measures. Mayo Clin Proc 2002; 77:371–383.
3 Herrera FJ, Wong J, Chung F. A systematic review of postoperative recovery outcomes measurements after ambulatory surgery. Anesth Analg 2007; 105:63–69.
4 Anthony T, Hynan LS, Rosen D, et al. The association of pretreatment health-related quality of life with surgical complications for patients undergoing open surgical resection for colorectal cancer. Ann Surg 2003; 238:690–696.
5 Davenport DL, Henderson WG, Khuri SF, Mentzer RM Jr. Preoperative risk factors and surgical complexity are more predictive of costs than postoperative complications: a case study using the National Surgical Quality Improvement Program (NSQIP) database. Ann Surg 2005; 242:463–468, discussion 8–71.
6 Tang TY, Walsh SR, Fanshawe TR, et al. Comparison of risk-scoring methods in predicting the immediate outcome after elective open abdominal aortic aneurysm surgery. Eur J Vasc Endovasc Surg 2007; 34:505–513.
7 Veltkamp SC, Kemmeren JM, van der Graaf Y, et al. Prediction of serious complications in patients admitted to a surgical ward. Br J Surg 2002; 89:94–102.
8 Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patient-based outcome measures for use in clinical trials. Health Technol Assess 1998; 2:i–i10, 1–74.
9 Ware JE Jr, Manning WG Jr, Duan N, et al. Health status and the use of outpatient mental health services. Am Psychol 1984; 39:1090–1100.
10 Fung D, Cohen MM. Measuring patient satisfaction with anesthesia care: a review of current methodology. Anesth Analg 1998; 87:1089–1098.
11 Orkin FK, Cohen MM, Duncan PG. The quest for meaningful outcomes. Anesthesiology 1993; 78:417–422.
12 Gandek B, Ware JE Jr, Aaronson NK, et al. Tests of data quality, scaling assumptions, and reliability of the SF-36 in eleven countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 1998; 51:1149–1158.
13 Heyland DK, Hopman W, Coo H, et al. Long-term health-related quality of life in survivors of sepsis. Short Form 36: a valid and reliable measure of health-related quality of life. Crit Care Med 2000; 28:3599–3605.
14 Ware JE Jr, Gandek B, Kosinski M, et al. The equivalence of SF-36 summary health scores estimated using standard and country-specific algorithms in 10 countries: results from the IQOLA Project. International Quality of Life Assessment. J Clin Epidemiol 1998; 51:1167–1170.
15 Carli F, Mayo N, Klubien K, et al. Epidural analgesia enhances functional exercise capacity and health-related quality of life after colonic surgery: results of a randomized trial. Anesthesiology 2002; 97:540–549.
16 Myles PS, Reeves MD, Anderson H, Weeks AM. Measurement of quality of recovery in 5672 patients after anaesthesia and surgery. Anaesth Intensive Care 2000; 28:276–280.
17 Myles PS, Hunt JO, Nightingale CE, et al. Development and psychometric testing of a quality of recovery score after general anesthesia and surgery in adults. Anesth Analg 1999; 88:83–90.
18 McHorney CA, Ware JE Jr, Raczek AE. The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 1993; 31:247–263.
19 Ware JE Jr, Gandek B. Overview of the SF-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project. J Clin Epidemiol 1998; 51:903–912.
20 Lee A, Lum ME. Measuring anaesthetic outcomes. Anaesth Intensive Care 1996; 24:685–693.
21 Garratt A, Schmidt L, Mackintosh A, Fitzpatrick R. Quality of life measurement: Rliographic study of patient assessed health outcome measures. BMJ 2002; 324:1417.
22 Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I: Conceptual framework and item selection. Med Care 1992; 30:473–483.
23 Myles PS, Hunt JO, Fletcher H, et al. Relation between quality of recovery in hospital and quality of life at 3 months after cardiac surgery. Anesthesiology 2001; 95:862–867.
24 Eberhart LH, Greiner S, Geldner G, Wulf H. Patient evaluation of postoperative recovery. An evaluation of the QoR scores in 577 patients. Anaesthesist 2002; 51:463–466.
25 Guyatt GH, Townsend M, Berman LB, Keller JL. A comparison of Likert and visual analogue scales for measuring change in function. J Chronic Dis 1987; 40:1129–1133.
26 Wu CC, Ho WM, Cheng SB, et al. Perioperative parenteral tranexamic acid in liver tumor resection: a prospective randomized trial toward a ‘blood transfusion’-free hepatectomy. Ann Surg 2006; 243:173–180.
27 Lang M, Niskanen M, Miettinen P, et al. Outcome and resource utilization in gastroenterological surgery. Br J Surg 2001; 88:1006–1014.
28 Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240:205–213.
29 Cohen J. Statistical power analysis for the behavorial sciences. 2nd edn. Mahwah, NJ: Lawrence Erlbaum; 1988.
30 Houle TT. Importance of effect sizes for the accumulation of knowledge. Anesthesiology 2007; 106:415–417.
31 Ware JE. Measuring patients' views: the optimum outcome measure. BMJ 1993; 306:1429–1430.
32 Weinberger M, Oddone EZ, Samsa GP, Landsman PB. Are health-related quality-of-life measures affected by the mode of administration? J Clin Epidemiol 1996; 49:135–140.
33 Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002; 236:397–406, discussion 407.
34 Maier-Riehle B, Zwingmann C. Effect strength variation in the single group prepost study design: a critical review. Rehabilitation (Stuttg) 2000; 39:189–199.

anaesthesia; epidural; general; hepatectomy; postoperative complications; recovery of function

© 2009 European Society of Anaesthesiology