Analgesia-related adverse effects and complications remained rare throughout the observation period (Table 7). Sedation (patient difficult to awake or respiratory frequency <8 min−1) was reported in 0.2% of patients.
The pain team met every 3-5 months with the head nurses and surgeons of all wards, and a representative of the hospital pharmacy. During these meetings, results were presented and discussed. A feedback example is presented in Figure 3. In the case of obvious deficits in process quality (i.e. decrease in the number of prescription sheets), precise actions were recommended to all involved parties (for instance, the ward nurse shall not accept the patient without this sheet). In the case of outcome impairment, changes in results were visualised and possible causes (i.e. change in pain management, staff fluctuation) were discussed. The results of these discussions were disseminated in written form.
Surveys show considerable deficits in postoperative pain management throughout the world [2,3]. In an investigation of postoperative pain services in Europe, a few only reported on routine pain assessment and documentation, and occasional outcome measurements were performed in 40% of the hospitals only . Few studies have reported on pre-post-improvement of acute pain management by single- or short-time interventions [11-13]. We report on the implementation of a CQI programme in postoperative pain management that has been performed over the past 5 yr and that included more than 6500 patients. Cornerstones of this project were continuous quality assessments, regular benchmarking and implementation of feedback mechanisms. Process as well as outcome quality parameters improved over time, partly reaching statistical significance.
We were able to show an improvement in pain on ambulation from 3.9 ± 2.3 to 3.6 ± 2.3 on a 11-point NRS over 5 yr. Maximal pain and pain at rest showed no change over time. In general, pain intensity was at least equal [4,13] or below [2,14,15] those published in recent reports of postoperative pain.
One could argue about the clinical relevance of such an improvement. Several projects studying one-time interventions showed more rapid changes in outcome parameters. In our preceding project, pain intensity decreased by 10% after 1 yr. However, the degree of improvement depends on the situation at the beginning of an intervention, allowing more progress when starting from a worse level of quality. Also, it seems to be very difficult to maintain such results over a longer period . In many short-lasting interventions, success might be triggered not specifically by the CQI activity but by the psychological stimulus of healthcare providers who are pleased to receive attention as long as the activity is in progress but who are losing interest as soon as the project has finished (‘Hawthorne effect’) . Thus, we are convinced that any measure that helps to assure a continuous level of quality for an extended period of time is of crucial importance. To our knowledge, there are no reports to date about similar CQI projects over a time period of several years in the pain setting.
In the field of quality improvement, recommended techniques have progressed from simple quality control to quality management consisting of continuous analysis, intervention and re-evaluation . The plan-do-check-act (PDCA) cycle is a well-known model for continuous process improvement . Continuous data analysis, routine feedback and benchmarking processes provide several advantages. First, the problem of influencing results by the evaluation itself can be reduced. Second, participants are encouraged to improve outcome not only for a limited time. Third, staff members become accustomed to this method of quality improvement, and, ideally, accept it as a means to develop good medical practice and clinical excellence.
Another advantage of continuous data assessment should be mentioned. Usually, benchmarking processes compare different competitors. In the beginning of our project, we also used this technique. However, obvious differences with other wards were often excused by co-factors that are difficult to adjust as, for instance, type of surgery, differences in patient demography or co-morbidity. Acceptance of feedback processes rose substantially when we started to compare wards not only with each other but with their own past performance; it was their own performance which now appeared to be the benchmark.
Outcome and process quality parameters
In a survey of more than 200 US institutions, more than 90% reported to have a continuos improvement programme in pain management . On a closer look, however, many respondents did not collect outcome data as pain intensity or impact of pain on functioning. The authors underlined the overwhelming importance of assessing parameters of outcome in any CQI activities .
We focused on three parameters of pain intensity as outcome indicators. However, methodology of pain measurement is controversial [20,21]. We observed that even in patients with low pain intensity at rest, high levels of maximal pain or of pain on ambulation may occur. Apparently maximal pain and/or pain on ambulation are more sensitive indicators for insufficient pain management than pain at rest or current pain. It is debatable as to whether ‘absolute’ benchmarks (i.e. ‘pain intensity < 4 on the 11-point numerical scale’) are appropriate targets of acute pain management. ‘Satisfactory’ pain intensity levels will differ according to the extent of surgery and probably many other factors. A NRS of 3 may be adequate after appendectomy but too high after minor surgery. Moreover, studies that recommend a fixed pain intensity level as a target of pain therapy may not be transferable to other patient groups. Also, in many of these recommendations, it is not specified whether an approved pain level (i.e. 3 or 4 on the 11-point numerical scale) refers to acute, average or maximal pain at rest or on movement.
Outcome questions describing how pain interferes with important activities like deep breathing, mobilisation, or sleep might also be relevant clinical indicators for the quality of pain management [6,22] and could be used to define outcome benchmarks. Dietrick-Gallagher and co-workers suggested that a given number of patients (i.e. 90%) meeting functional outcome criteria might be a meaningful target in postoperative pain therapy . However, standardised tools for the evaluation of functional aspects in postoperative pain control have not been evaluated yet.
Although improvement of outcome is the main purpose of CQI, comprising parameters of process quality may help to identify obstructive factors . In one of the earliest studies on quality improvement in acute pain management, it was shown that introduction of the measurement of pain in itself improved pain intensity significantly . We decided to incorporate two questions which may indicate possible inadequacies of the different team members who are involved in pain management, i.e. the number of missing prescriptions and assessment sheets at bedside; these may reflect deficits in the communication between anaesthetist, recovery room and ward. The number of patients not receiving their routine non-opioid analgesic may indicate shortcomings in prescription or nursing care. Parameters of process quality as well as structural quality should be determined according to their likely impact on outcome . Unfortunately, process quality is often considered as an important value in itself. In contrast, what is important for the patient is not how but if successful and safe pain treatment is provided.
Our patients reported a remarkable low incidence of adverse effects like nausea, vomiting or sedation. This may partly be explained by the routine use of non-opioid analgesics which are part of our medical concept and that may reduce opioid requirements. Also, on the first postoperative day, most anaesthesia-related adverse effects have disappeared. However, it is possible that an earlier and more frequent survey would have shown a higher incidence of these complications.
It would be desirable to collect more information on adverse reactions of non-opioid analgesics as gastro-intestinal, renal or haematological adverse effects. However, most of these complications will develop later than during the first postoperative day, and assessment would require some laboratory or other exploration.
Our report is subject to several limitations. The method of randomization does not meet the requirements of a prospective, clinical study; however it seems to be an acceptable compromise for a CQI project performed in clinical routine practice. Patients who had surgery on a Friday were not equally represented since on weekends, no data collection could be performed. In our institution, different types of surgery are equally spread over all weekdays.
A critical issue of any survey of subjective parameters is the potential interviewer-patient interaction bias. Complete blinding of the interviewer was not feasible; therefore, we did not rely on data obtained by the staff of the wards, we separated assessments from clinical routine, and we undertook great efforts to train the interviewer for data collection.
Another possible limitation of our study was that hard outcome criteria, such as length of hospital stay or morbidity, were not included. We studied a wide range of surgical procedures including a number of minor operations where consequences on length of hospital stay or morbidity were unlikely. Also, improvement in pain control itself should be valued as an ethical goal of medical treatment, and discussion should not be limited to economic aspects.
It would have been highly desirable to implement more process and outcome parameters as well as possible complications and adverse effects. However, a CQI project that works under routine conditions must focus on a limited number of data. In our project, these parameters had to meet the following criteria: displaying outcome quality of pain therapy from the patient's perspective, indicating processes only that were thought to be important and being easily available. But even if much more possible co-factors would have been obtained it is questionable whether a continuous data collection like ours is suitable for discovering causal relationships between interventions and possible complications because there is no randomisation of treatment, and it is not possible to control all variables. This is a common problem in all performance measurement studies [13,26]. It should be emphasized that quality improvement procedures, as demonstrated in this study, must be framed by appropriate measures for feedback and ‘management of change’. This is probably an even more important part of a CQI programme than to establish data collecting and feedback.
Meanwhile, an extension of this project to other hospitals is supported by a grant from the German Federal Ministry of Health and Social Security. Outcome parameters were expanded by assessment of functional impact of pain, and some more clinical parameters (i.e. technique of pain management, type and duration of surgery, method of anaesthesia) were included. A web-based benchmark software was developed which allows immediate data analysis and feedback showing each participant his/her results compared with all others via a secured internet access.
In conclusion, a CQI process was established and has been successfully used in clinical routine for more than 5 yr. Some parameters of process and outcome quality improved over the observation period, others remained constant. Transparency and multidisciplinary teamwork as well as benchmarking and prompt feedback mechanisms seem to be key elements of the successful implementation of a quality management initiative.
This project was supported by the staff of the Departments of Visceral Surgery, Orthopaedic/Trauma Surgery and Cardiothoracic Surgery. We thank Sigrid Oehler, Head of the Surgical Nursing Team, for her special assistance and permanent support. The extension of this project is supported by a grant from the German Federal Ministry of Health and Social Security (217-43794-6/3).
1. Kehlet H. Effect of postoperative pain treatment on outcome-current status and future strategies. Langenbecks Arch Surg
2. Apfelbaum JL, Chen C, Mehta SS, Gan TJ. Postoperative pain experience: results from a national survey suggest postoperative pain continues to be undermanaged. Anesth Analg
: 534-540, table of contents.
3. Rawal N. Acute pain services revisited- good from far, far from good? Reg Anesth Pain Med
4. Dolin SJ, Cashman JN, Bland JM. Effectiveness of acute postoperative pain management: I. Evidence from published data. Br J Anaesth
5. Rawal N, Berggren L. Organization of acute pain services: a low-cost model. Pain
6. American Pain Society Quality of Care Committee. Quality improvement guidelines for the treatment of acute pain and cancer pain. JAMA
7. Lack JA, White LA, Thoms GM, Rollin AM. Raising the Standard: The Royal College of Anaesthetists, 2000.
8. Allcock N. The use of different research methodologies to evaluate the effectiveness of programmes to improve the care of patients in postoperative pain. J Adv Nurs
9. Meissner W, Ullrich K, Zwacka S, Schreiber T, Reinhart K. [Quality management in postoperative pain therapy.] Anaesthesist
10. Rawal N, Allvin R. Acute pain services in Europe: a 17-nation survey of 105 hospitals. The EuroPain Acute Pain Working Party. Eur J Anaesthesiol
11. Coleman SA, Booker-Milburn J. Audit of postoperative pain control. Influence of a dedicated acute pain nurse. Anaesthesia
12. Tighe SQ, Bie JA, Nelson RA, Skues MA. The acute pain service: effective or expensive care? Anaesthesia
13. Bardiau FM, Taviaux NF, Albert A, Boogaerts JG, Stadler M. An intervention study to enhance postoperative pain management. Anesth Analg
14. McNeill JA, Sherwood GD, Starck PL, Thompson CJ. Assessing clinical outcomes: patient satisfaction with pain management. J Pain Symptom Manage
15. Miaskowski C, Nichols R, Brody R, Synold T. Assessment of patient satisfaction utilizing the American Pain Society's Quality Assurance
Standards on acute and cancer-related pain. J Pain Symptom Manage
16. Shayer M. Problems and issues in intervention studies. In: Demetriou A, Shayer M, Efklides A (eds). Neo-Piagetian Theories of Cognitive Development: Implications and Applications for Education
. London: Routledge, 1992: 107-121.
17. Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physician performance. A systematic review of the effect of continuing medical education strategies. JAMA
18. Deming WE. Out of the Crisis
. Cambridge, MA: MIT Press, 1986.
19. Ward S, Donovan M, Max MB. A survey of the nature and perceived impact of quality improvement activities in pain management. J Pain Symptom Manage
20. Aubrun F, Paqueron X, Langeron O, Coriat P, Riou B. What pain scales do nurses use in the postanaesthesia care unit? Eur J Anaesthesiol
21. Jensen MP, Chen C, Brugger AM. Postsurgical pain outcome assessment. Pain
22. Crowley B, Lubesnick K, Sylwestrak ML. Clinical indicators: a tool for improving pain management documentation. J Nurs Care Qual
23. Dietrick-Gallagher M, Polomano R, Carrick L. Pain as a quality management initiative. J Nurs Care Qual
24. Gould TH, Crosby DL, Harmer M et al
. Policy for controlling pain after surgery: effect of sequential changes in management. BMJ
25. Gordon DB, Pellino TA, Miaskowski C et al
. A 10-year review of quality improvement monitoring in pain management: recommendations for standardized outcome measures. Pain Manag Nurs
26. Joint Commission on Accreditation of Healthcare Organizations. Improving the Quality of Pain Management through Measurement and Action
. Oakbrook Terrace, IL: JCAHO, 2003.
Keywords:© 2006 European Society of Anaesthesiology
QUALITY ASSURANCE; health care; BENCHMARKING; PAIN POSTOPERATIVE