Pain measurement is mandatory in the postanaesthesia care unit (PACU) to assess pain intensity, to control the efficacy of analgesic treatment and to ensure a confident relationship between the patient and the medical team . Quality assurance programmes include development of pain assessment in the PACU and in the ward by nurses trained to use unidimensional pain scales . Self-assessment, using unidimensional methods that measure only the sensory component of the painful experience, has been validated in acute pain management ; it reduces the risk of under or over assessment of pain by nurses . Several scales have been proposed to assess pain . The visual analogue scale (VAS) is a handheld slide-rule type with two end-points, 0 labelled 'no pain' and 100 labelled 'the worst imaginable pain'; the cursor is placed between these end-points by the patient . The numerical rating scale consists of a series of numbers ranging from 0 'no pain' to 10 (or 100) 'the worst imaginable pain'; the patient chooses the number that best corresponds to the intensity of pain. The verbal rating scale consists of a series of verbal pain descriptors: the patient reads the list and chooses the one word that best describes the intensity of his or her pain; a score is assigned to each descriptor (e.g. no pain = 0, mild = 1, moderate = 2, severe = 3) . Finally, a simple behavioural scale can be used to decide if the patient experiences any pain . This simple and subjective scale has not been validated; it consists of a complaint behaviour (=pain treatment) and relief behaviour (=no treatment).
The VAS is considered the 'gold standard' method for the assessment of chronic and acute pain [3,7]. Nevertheless, in some clinical situations and in some patients it may be difficult, or even impossible, to use [7,8]. We performed a prospective observational study to assess the use of the VAS and other pain scales by nurses in the PACU. We studied the reasons that lead to the use of a method other than the VAS and assessed the opinions of nurses about pain scales. We also examined if some differences existed between young and elderly patients in the nurses' choice of pain scales.
For many years, the nurses of our PACU have been trained systematically to assess pain using, first, the VAS and, second, a 0-10 point numerical rating scale, a 0-100 point numerical rating scale or a 0-3 item visual rating scale [2,9]. Each nurse possesses a personal VAS ruler and can easily obtain a replacement if it is lost. If the patient is unable to use a unidimensional scale, the nurse is allowed to use a simple behavioural scale, despite its low correlation with the sensation of pain.
During the preoperative visit, which took place at least 48 h before surgery, patients were familiarized with the recording of the VAS as a method to assess and treat postoperative pain. The assessment of pain was used to decide and perform intravenous (i.v.) morphine titration, according to a strict protocol applied by the nurses in the PACU, and to decide and perform subcutaneous morphine titration thereafter [2,9,10]. After endotracheal extubation and when awake, patients were questioned about the presence of pain; they were also asked to rate the intensity of their pain. When patients experienced pain, i.v. morphine was titrated every 5 min by 3 mg increments (2 mg in patients weighing ≤60 kg) and pain was assessed every 5 min until pain relief. Morphine titration was not adapted to age . Absence of pain (i.e. pain relief) was defined as a VAS value <30 mm, a numerical rating scale value <30 or 3, respectively, or a verbal rating scale of 0 or 1.
Several prospective studies have been performed in our PACU in which only patients able to use a VAS or numerical rating scale have been included [2,9]. The present study was conducted over 16 months; during that period no other prospective study on pain was in progress. The nurses completed a form indicating the method they used to assess pain. When the VAS was not used, nurses were asked to indicate the reason why this recommended scale had not been used. There were nine possible choices: (1) another method was easier to use; (2) the patient was in too much pain to use the VAS; (3) there was a lack of understanding of the VAS by the patient; (4) appropriate communication with the patient was not possible; (5) personal choice of the nurse; (6) difficulties in manipulating the cursor by the patient; (7) lack of ruler; (8) another method was considered by the nurse to be faster; and (9) unspecified.
Patients were divided into two groups according to age: young patients (<70 yr) and elderly (≥70 yr). The method used to assess pain and the reasons why the VAS was not used were compared between these two groups.
Each nurse in the PACU also completed a simple questionnaire to rate his or her opinions, first on the simplest tool to use, then on the most reliable tool in assessing pain. This study did not modify the usual management of pain control in the PACU. All patients who needed morphine received i.v. morphine titration depending on the level of their pain intensity as described above.
Data are expressed as mean ± standard deviation (SD), and as numbers and percentages. Comparison between two means was performed using the t-test. Comparison between two percentages was performed using Fisher's exact method. All P values were two-tailed. P < 0.05 was considered as significant.
Six hundred patients admitted to the PACU after orthopaedic (n = 341, 57%), urological (84, 14%), abdominal (87, 14%), vascular (53, 9%), gynaecological (30, 5%) and other types of surgery (5, 1%) were included in the study. There were 320 (53%) males, and 280 (47%) females. The average age was 51 ± 17 yr (range 15-92 yr); 96 (16%) patients were elderly. The average dose of i.v. morphine titration in the PACU was 9.8 ± 6.5 mg (range 0-40 mg), or 0.14 ± 0.10 mg kg−1.
The VAS was most frequently used (53%), followed by the numerical rating scale (30%) and the verbal rating scale (12%). Only 5% of the patients could not have their pain assessed with an unidimensional scale. No significant differences between young and elderly patients were observed (Table 1).
When nurses did not use the VAS, the most frequently cited reasons (41%) were related to the fact they preferred other methods. In 56% of cases, the reason was related to the patient, mainly because the patient was in too much pain to use the VAS (22%). There were no significant differences between young and elderly patients (Table 2). When an easier method was preferred (n = 63), the numerical (0-10) rating scale was chosen in 60% and the verbal rating scale in 24% of cases. When the patient was in too much pain (n = 59), the numerical rating scale was chosen in 54% and the behavioural scale in 27% of cases. When the patient had problems in understanding and communicating (n = 43 and 42, respectively), the nurses preferred using the numerical (0-10) rating scale (49 and 40%, respectively). When nurses' personal choice was the main determinant, the numerical (0-10) rating scale proved the most popular tool (73%).
During the study period, 79 nurses filled in the questionnaire. The simplest tool was considered the numerical (0-10) rating scale; the most reliable tool was considered the VAS (Table 3).
The main findings of our study are that the nurses used the VAS in only 53% of pain assessments, even though this was the recommended tool; the numerical rating scale was preferred since it was more rapid and easier to use. No significant difference could be demonstrated between young and elderly patients.
Some authors have reported that 2% of patients are unable to use the numerical rating scale and 7-11% are unable to use the VAS or find it confusing [11,12]. In our study, 5% of patients were not able to use any of the unidimensional scales; their pain was assessed using the behavioural or the verbal rating scale. Although the VAS was the recommended tool in our institution, only 53% of the patients were assessed using this method since the nurses found other tools faster or easier to use (Table 2). DeLoach and colleagues  noticed that the VAS was difficult to use in the postoperative period due to residual anaesthesia, blurred vision or nausea. It should be pointed out that the most frequent reason why patients failed to use the VAS was the severity of pain. In our previous prospective studies [2,9], when pain was too severe to obtain a VAS, it was arbitrarily scored 100, and pain assessment using the VAS was usually possible after partial pain relief had been obtained. Lack of communication and understanding are also important reasons for preventing the use of the VAS in the PACU (Table 2). These results concord with those of Jensen and colleagues  who demonstrated that the VAS was more difficult to employ than the numerical rating scale or an adjective categorical scale in patients with chronic pain. In the same way, Berthier and colleagues  performed a comparative study on methods of measuring acute pain intensity in an emergency department. The VAS could not be used in 19.5% of patients with trauma and in 4.5% of non-trauma patients. Failure was mainly due to motor incapacity to manipulate the cursor, cognitive functional disorders, visual disorders and language incomprehension. In our study, the personal choice of the nurses was the major reason for not using the VAS. Using this scale requires the assistance of a nurse and is time consuming. Nurses are often hard pressed in the PACU (Table 2). In contrast, nurses thought the VAS was the most reliable method to assess pain, but their training might have influenced them since the VAS was the recommended method to assess pain in our PACU.
The numerical rating scales are simple to use for nurses and easy to understand for patients. These tools eliminate the need for the visual and motor coordination required to complete the VAS and thus are more likely to be completed. In patients with chronic pain, Jensen and colleagues  demonstrated that the numerical rating scale appeared to be the most practical index. In the postoperative period, DeLoach and colleagues  appraised that the numerical rating scale could overcome most of the VAS-related difficulties, and demonstrated that numerical rating scales correlated well with the VAS. Berthier and colleagues  showed that the numerical scale and VAS were closely correlated for trauma and non-trauma patients in an emergency department. Moreover, in this study, the numerical rating scale was successfully used by 96% of patients versus 85% who used the VAS. In an acute pain model (after oral surgery), Breivnik and colleagues  reported that the sensitivity of the VAS and numerical rating scale were approximately equal. Conversely, in our recent prospective studies [2,5], these two methods were used alternatively, and their measures were considered equivalent for the assessment of analgesia with intravenous morphine.
Behavioural scales are particularly useful for patients who cannot answer to or understand self-reporting methods. This method has not been validated for the postoperative period. Often nurses are self-trained to observe facial expressions, to hear complaints and to decide whether or not they shall initiate i.v. morphine titration according to a given protocol. One of the main disadvantages of the behavioural methods is that they do not quantify pain. In our study, we used a very simple scale (pain/no pain). Although our nurses were asked to use alternative methods, the behavioural scale was used by 5% of the patients only. The problem of pain evaluation in patients who are not able to use self-reporting methods was raised by Terai and colleagues  in the intensive care unit. They found that the observer-reported faces scale - in comparison with self-reported VAS - was useful for pain evaluation in this situation, although this tool was not always reliable in case of slight or moderate pain. There is obviously a need for a validated tool for patients in the postoperative period.
Pain complaints have been described as being less frequent in the elderly . Inadequate postoperative analgesia is an important problem in the elderly who are reported to consume less opioids [18,19]. Frequently, lack of information about pain, lack of pain assessment and fear of increased risks of adverse effects lead to under treatment of pain in the elderly. In the present study, there was no significant difference between the young and the elderly when considering the method used to assess pain (Table 1), or when considering the reasons why the VAS was not used (Table 2). These results are not in accordance with some previous studies. Elderly patients have been reported to experience more difficulties with the VAS . Using an acute pain model, Berthier and colleagues  showed that patients aged more than 65 yr were more often than younger patients unable to give pain scale responses (36 versus 14%, P < 0.001). Success was less frequent with the VAS compared with the numerical or verbal rating scales. In a chronic pain model, Jensen and colleagues  observed there was a significant correlation between age and incorrect response to the VAS. This method is more difficult to use than the numerical method or an adjective categorical scale [3,11]. However, most of these studies were performed in patients with chronic or perioperative pain but not for the immediate postoperative pain assessment when morphine titration is required because of severe pain [3,20]. Our study suggests that elderly patients should be assessed as young patients, as previously demonstrated for i.v. morphine administration .
Our study has some limitations. First, it shows what nurses do for routine care; it does not show what should be done. For instance, 17% of the patients were assessed using non-validated methods such as a behavioural scale. Further studies should be directed to the validation of methods that are required when the VAS and/or numerical rating scale are not feasible. Second, our study was conducted in a PACU where nurses have been trained for years to assess pain using various tools. The fact that trained nurses were using the VAS may have been expected to favour this method. Third, we actually did not compare these methods for pain assessment or pain control. Fourth, we did not assess what was performed thereafter in the surgical wards, and thus our results apply to the immediate postoperative period in the PACU only. Lastly, we did not study the very elderly population (>90 yr) who may behave differently and/or induce a different behaviour from the nurses.
In conclusion, we demonstrated that the nurses working in our PACU used the VAS in only 53% of the patients. The numerical rating scale was the preferred tool and there was no difference between young and elderly patients in the choice of the method or the reasons for not using the VAS. No pain scale has ever been validated in the PACU. It is necessary to perform studies to validate easy, fast and powerful methods like the numerical rating scale or a behavioural scale.
We thank Dr David Baker, FRCA (Département d'Anesthésie-Réanimation chirurgicale, CHU Necker-Enfants Malades, Paris, France) for reviewing the manuscript.
1. Feeley TW. The postanesthesia care unit. In: Miller RD, ed. Anesthesia,
3rd edn. New York, USA: Churchill Livingstone, 1990: vol. 3, 2113-2133.
2. Aubrun F, Monsel S, Langeron O, Coriat P, Riou B. Postoperative titration of intravenous morphine. Eur J Anaesthesiol
3. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain
4. Zalon ML. Nurses' assessment of postoperative patients' pain. Pain
5. Huskisson EC. Measurement of pain. Lancet
6. Katz J, Melzack R. Measurement of pain. Surg Clin North Am
7. Conférence de Consensus. Prise en charge de la douleur postopératoire chez l'adulte et l'enfant. Ann Fr Anesth Réanim
8. Ricard-Hibon A, Chollet C, Saada S, Loridant B, Marty J. A quality control program for acute pain management in out-of-hospital critical care medicine. Ann Emerg Med
9. Aubrun F, Monsel S, Langeron O, Coriat P, Riou B. Postoperative titration of intravenous morphine in the elderly. Anesthesiology
10. Aubrun F, Kalfon F, Mottet P, Langeron O, Riou B, Coriat P. Adjuntive analgesia with intravenous propacetamol does not reduce morphine-related adverse effects. Br J Anaesth
11. Kremer E, Atkinson JH, Ignelzi RJ. Measurement of pain: patient preference does not confound pain measurement. Pain
12. Revill SI, Robinson JO, Rosen M, Hogg MI. The reliability of a linear analogue for evaluating pain. Anaesthesia
13. DeLoach LJ, Higgins MS, Caplan AB, Stigg JL. The visual analogue scale in the immediate postoperative period: intrasubject variability and correlation with a numeric scale. Anesth Analg
14. Berthier F, Potel G, Leconte P, Touze MD, Baron D. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med
15. Breivnik EK, Björnsson GA, Skovlund E. A comparison of pain rating scales by sampling from clinical trial data. Clin J Pain
16. Terai T, Yukioka H, Asada A. Pain evaluation in the intensive care unit. Observer reported faces scale compared with self reported visual analog scale. Reg Anesth Pain Med
17. Wall RT. Use of analgesics in the elderly. Clin Geriatr Med
18. Portenoy RK, Kanner RM. Patterns of analgesic prescription and consumption in a university-affiliated community hospital. Arch Intern Med
19. Faherty BS, Grier MR. Analgesic medication for elderly people post-surgery. Nurse Res
20. Thomas T, Robinson C, Campion D, McKell M, Pell M. Prediction and assessment of the severity of postoperative pain and of satisfaction with management. Pain