Advertisement

Can severe acute pain escape visual analog scale screening in the ED?

      To the Editor:—Pain relief is certainly one of the most important components of medical care for all patients of all times throughout the world. In EDs, pain prevalence is very high; it was estimated to be over 80% in several French multicenter studies.
      • Tempelhoff C
      • Tempelhoff G
      La douleur dans les services d’accueil et d’urgence état des lieux, étude multicentrique.
      ,
      • Simon N
      • Alland M
      • Brun-Ney D
      • et al.
      Le traitement médicamenteux de la douleur de l’adulte dans un service d’accueil et d’urgence.
      It is now widely admitted that pain treatment does not hinder diagnostic investigations or therapeutic procedures.
      • Pace F
      • Burke T.F
      Intravenous morphine for early pain relief in patients with acute abdominal pain.
      ,
      • Brasseur L
      • Larue F
      • Charpak Y
      L’analgésie postopératoire en France. Etude sur l’attitude des médecins prescripteurs.
      ,
      • Boureau F
      • Doubrère J.F
      Le concept de douleur. Du symptome au syndrome.
      Therefore, appropriate analgesia should be a priority in EM.
      • Cordell W.H
      Pain—control research opportunities and future directions.
      However, pain control practices are often inadequate in ED for several reasons.
      • Simon N
      • Alland M
      • Brun-Ney D
      • et al.
      Le traitement médicamenteux de la douleur de l’adulte dans un service d’accueil et d’urgence.
      ,
      • Van der Roost D
      Intervention en soins infirmiers. Face à la douleur aux urgences: action.
      ,
      • Berthier F
      • Le Conte P
      • Garrec F
      • Potel G
      • Baron D
      Analyse de la prise en charge de la douleur aiguë dans un service d’accueil et d’urgence.
      The very first of these reasons (at least chronologically if not in general) is the underestimation of patients’ suffering. One-dimensionalscales such as the Visual Analog Scale (VAS) and Verbal Rating Scale (VRS) have been proposed for routine pain intensity evaluation in post-operative and emergency situations.
      • Boureau F
      • Luu M
      • Héricord P
      Evaluation de l’intensité de la douleur au service d’accueil et des urgences, intér.
      ,
      • Hirsowski F
      • Boureau F
      Nécessité d’évaluer la douleur en pratique quotidienne.
      In 1994, a Canadian Association for Emergency Physicians consensus recommended VAS for initial pain and pain relief evaluation.
      • Ducharme J
      Emergency pain management a Canadian Association of Emergency Physicians consensus document.
      However, although VAS is the actual standard used in most ED clinical trials dealing with pain management,
      • Hirsowski F
      • Boureau F
      Nécessité d’évaluer la douleur en pratique quotidienne.
      ,
      • Yealy D.M
      Prehospital pain research.
      it is not actually systematically used by emergency practitioners to evaluate pain intensity,
      • Simon N
      • Alland M
      • Brun-Ney D
      • et al.
      Le traitement médicamenteux de la douleur de l’adulte dans un service d’accueil et d’urgence.
      ,
      • Berthier F
      • Le Conte P
      • Garrec F
      • Potel G
      • Baron D
      Analyse de la prise en charge de la douleur aiguë dans un service d’accueil et d’urgence.
      ,
      • Berthier F
      • Potel G
      • Leconte P
      Comparative study of methods of measuring acute pain intensity in a ED.
      maybe because its pertinence to reveal severe acute pain (SAP) in ED lacks evidence. The purpose of the study was to test VAS ability to identify SAP, to establish the VAS level best corresponding to SAP, to measure its sensitivity, specificity, positive and negative predictive value (PPV and NPV) and to describe the characteristics of SAP that escape to VAS screening in the ED.
      During 7 consecutive days, patients aged over 15 years presenting to the ED of any of the six hospitals of the district were enrolled in the study. Criteria for excluding patients were life-threatening distress, refusal to participate in the study, intellectual functional alteration (such as alcoholism or senile dementia), and inability to understand French.
      Pain intensity at time of arrival to the ED was assessed with the VRS and VAS. Questionnaires using a six-point box VRS to collect VRS level were blindly paired to medical forms recording VAS level. VRS and VAS documents with a common anonymous number were separated at the time of the patient’s arrival to the ED. VRS questionnaires were completed by patients and kept aside from the emergency practitioner’s view while he filled out the medical form, including VAS level. The other data collected on these documents were patient’s satisfaction concerning pain management (satisfied vs. not satisfied), VRS final level (six-point box VRS scale), pain origin (trauma vs. nontrauma), pain type (continuous vs. intermittent), and pain treatment (analgesia in ED vs. no analgesia in the ED). The anonymous numeration authorized paired analysis of blindly collected data.
      SAP was defined by patient’s initial VRS score = severe pain (severe, intense, or unbearable pain). The ability of the VAS to identify SAP in the ED was measured with a VAS score > 30 mm, a VAS score > 40 mm, and a VAS score > 50 mm as screening tests. MacNemar tests, Cohen’s concordance, and Youden’s coefficients were analyzed to find out which VAS level best corresponded to SAP definition.
      Of the 2929 patients aged over 15 years presenting to the ED during the survey, 1477 were excluded: 232 for life-threatening distress, 332 for refusal to participate in the study, 297 for intellectual alteration, and 616 for inability to understand French. A total of 1064 double-paired fulfilled forms were analyzed for the 1452 included ED patients. Mean simultaneous response rate was 72%.
      FIGURE 1, FIGURE 2 show VRS and VAS pain-level distribution of ED patients. Pain prevalence in ED was estimated 84% with the VRS and 75% with the VAS (P < 0.01). SAP prevalence was estimated 36% with the VRS, 42% with the VAS > 30 mm (P < .01), 37% with a VAS > 40 mm (P > .05), and 32% with a VAS > 50 mm (P < .01).
      Figure thumbnail GR1
      FIGURE 1Verbal Rating Scale pain level distribution of ED patients.
      Figure thumbnail GR2
      FIGURE 2Visual Analog Scale pain level distribution of ED patients.
      Table 1 shows the results of the MacNemar test, chi-squared prevalence comparison, Cohen’s concordance, and Youden’s coefficients analysis as well as sensitivity, specificity, PPV and NPV for VAS testing. VAS > 40 mm appears to be the best level to define SAP in the ED (P > .05). VAS > 30 mm leads to overestimating SAP prevalence and VAS > 50 mm leads to underestimating SAP prevalence (P < .01). The characteristics of VAS > 40 mm used as a screening test to identify SAP in ED are as follows: sensitivity 0.92, specificity 0.94, PPV 0.89, and NPV 0.95.
      TABLE 1Visual Analog Scale (VAS) Discriminant Power to Detect Severe Acute Pain in the ED
      VAS > 30 mmVAS > 40 mmVAS > 50 mm
      MacNemar paired analysisP < .01P > .05P < .01
      Chi-squared prevalence analysisP < .01P > .05P < .01
      Cohen’s concordance0.810.850.75
      Youden’s coefficient0.830.860.73
      Sensitivity0.930.920.78
      Specificity0.900.940.95
      Positive predictive value0.840.890.90
      Negative predictive value0.960.950.88
      Of 383 ED patients who declared VRS pain level = severe (SAP subgroup), VAS screening identified 351 scores > 40 mm. The occurrence of SAP escaping to VAS screening was 8% (CI95, 5–11%) among SAP group and 3% (CI95, 2–5%) among all ED patients. The 32 cases of SAP that escaped to VAS screening were compared with the 351 detected cases (Table 2) and appeared to be significantly different (P < .01): no analgesia administered during the ED stay (100%), spontaneous pain relief (75%), and intermittent type (44%).
      TABLE 2Characteristics of Severe Pain That Escape Visual Analog Scale (VAS) Screening
      VAS score > 4/10VAS score = 4/10P
      No35132P < .01
      Trauma/nontrauma56%/44%58%/42%P > .05
      Intermittent/continuous10%/90%44%/56%P < .01
      Spontaneous relief/persistent33%/67%75%/25%P < .01
      Analgesia in ED/no analgesia77%/23%0%/100%P < .01
      Satisfied patients/unsatisfied40%/60%38%/62%P > .05
      The exclusion criteria are those of similar designs.
      • Tempelhoff C
      • Tempelhoff G
      La douleur dans les services d’accueil et d’urgence état des lieux, étude multicentrique.
      ,
      • Berthier F
      • Le Conte P
      • Garrec F
      • Potel G
      • Baron D
      Analyse de la prise en charge de la douleur aiguë dans un service d’accueil et d’urgence.
      ,
      • Berthier F
      • Potel G
      • Leconte P
      Comparative study of methods of measuring acute pain intensity in a ED.
      The inability to understand written French language was the principle exclusion criteria of the study (21% of ED patients). The number of excluded patients is prone to selection bias but seems inevitable and is difficult to appreciate because it does not appear in all publications.
      • Tempelhoff C
      • Tempelhoff G
      La douleur dans les services d’accueil et d’urgence état des lieux, étude multicentrique.
      ,
      • Berthier F
      • Le Conte P
      • Garrec F
      • Potel G
      • Baron D
      Analyse de la prise en charge de la douleur aiguë dans un service d’accueil et d’urgence.
      ,
      • Yealy D.M
      Prehospital pain research.
      ,
      • Huskisson E.C
      Measurement of pain.
      A multicenter prospective survey requiring comparative and blindly paired data is a challenge given the constraints of ED in France. Considering that VRS and VAS documents were to be separately completed and later assembled for paired analysis, 72% response rate (1064 double-fulfilled forms of 1452 included ED patients) is a success.
      VRS and VAS are both considered as one-dimensional, simple, and validated methods for pain intensity evaluation.
      • Boureau F
      • Luu M
      • Héricord P
      Evaluation de l’intensité de la douleur au service d’accueil et des urgences, intér.
      ,
      • Berthier F
      • Potel G
      • Leconte P
      Comparative study of methods of measuring acute pain intensity in a ED.
      The only true experts in the matter are the patients themselves.
      • Berthier F
      • Potel G
      • Leconte P
      Comparative study of methods of measuring acute pain intensity in a ED.
      ,
      • Huskisson E.C
      Measurement of pain.
      In this study, VRS scoring was performed by ED patients alone completing a multiple-choice questionnaire. VRS was a completely self-rating procedure and therefore considered as the reference for SAP definition. VAS was performed by patients like in routine practice with the help of EPs who were expected to deliver the following message: Move the cursor along the blue line to indicate the intensity of your pain. The left end represents no pain and the right end represents the most intense and unbearable imaginable pain. The participation of the ED practitioner is important in the evaluation procedure, and VAS scoring cannot be considered as a completely self-rating method.
      • Boureau F
      • Luu M
      • Héricord P
      Evaluation de l’intensité de la douleur au service d’accueil et des urgences, intér.
      ,
      • Boureau F
      • Luu M
      • Gay C
      Les échelles d’évaluation de la symptomatologie douloureuse chronique.
      Naturally, VRS can only be a self-rating reference if the patient reads the pain evaluation form and uses a pen or pencil to tick the blanks, without the assistance of the ED team. That is why there are so many excluded patients (50%), and why the VRS written questionnaire cannot be recommended as a standard for pain evaluation in EM. On the other hand, the VAS procedure is easily accepted by ED patients, and pain scoring satisfies emergency practitioners. The question of VAS ability and discriminant power to identify SAP in the ED seems therefore justified and was the main purpose of this study.
      The difference between pain prevalence in the ED estimated with VRS (84%) or VAS (75%) demonstrates that ED practitioners using VAS tend to ignore a significant part of a patient’s suffering. The comparison of VRS and VAS pain-level distribution for ED patients (Fig. 2) shows that the underestimation of pain prevalence is the result of mild pain ignorance, as if mild pain was insufficient to activate a move of the VAS cursor.
      Although VAS is widely used in pain management trials and clinical practice, the answer to the question, “What VAS level best corresponds to SAP?” is not easy. The usual frontier separating light or moderate from important or severe pain varies from VAS > 30 mm to VAS > 50 mm.
      • Tempelhoff C
      • Tempelhoff G
      La douleur dans les services d’accueil et d’urgence état des lieux, étude multicentrique.
      ,
      • Van der Roost D
      Intervention en soins infirmiers. Face à la douleur aux urgences: action.
      ,
      • Ricard-Hibon A
      • Leroy N
      • Magne M
      • Leberre A
      • Chollet C
      • Marty J
      Evaluation de la douleur aiguë en médecine préhospitalière.
      Therefore, the discriminant power of the VAS as a screening test to identify SAP had to be analyzed considering 30 mm, 40 mm, and 50 mm as possible border levels. VAS > 40 mm appears to be the best screening test throughout all outcome measures. MacNemar (based on discordant paired data analysis) and chi-squared test (based on prevalence comparison) both conclude that VAS > 40 mm and the patient’s opinion through VRS are concordant to define SAP in the ED (P > .05), whereas VAS > 30 mm and VAS > 50 mm overestimate or underestimate SAP (P < .01). The other features of VAS screening tests (Table 1) lead to the same conclusion, and indicate that VAS > 30 mm is better than VAS > 50 mm. (We did not analyze other border levels such as 35 or 45 mm).
      Therefore, VAS is efficient for SAP identification. The high sensitivity (0.92) and NPV (0.95) of VAS level > 40 mm as a screening test prove that very few cases of SAP escape to VAS identification. Moreover, cases of SAP that do escape VAS screening in the ED do not significantly affect patient’s satisfaction despite the lack of analgesia, probably because of the very high frequency of spontaneous pain relief encountered in this particular subgroup (Table 2). Considering that light and momentary pains are not a priority for emergency practitioners, VAS is very pertinent in the ED.
      The best border level to define SAP with VAS in ED is VAS > 40 mm. Light pain is sometimes ignored, but very few cases of SAP escape VAS screening in the ED. Furthermore, undetected pains are ephemeral and do not affect a patient’s satisfaction. The high sensitivity and pertinence of the sliding cursor moving along the blue line should encourage the VAS for pain level scoring in EM.

      References

        • Tempelhoff C
        • Tempelhoff G
        La douleur dans les services d’accueil et d’urgence.
        Réan Urg. 1993; 2: 328-330
        • Simon N
        • Alland M
        • Brun-Ney D
        • et al.
        Le traitement médicamenteux de la douleur de l’adulte dans un service d’accueil et d’urgence.
        Réan Urg. 1993; 2: 321-327
        • Pace F
        • Burke T.F
        Intravenous morphine for early pain relief in patients with acute abdominal pain.
        Acad Emerg Med. 1996; 12: 1086-1092
        • Brasseur L
        • Larue F
        • Charpak Y
        L’analgésie postopératoire en France. Etude sur l’attitude des médecins prescripteurs.
        La presse médicale. 1992; 21
        • Boureau F
        • Doubrère J.F
        Le concept de douleur. Du symptome au syndrome.
        Doul et Analg. 1988; 1: 11-17
        • Cordell W.H
        Pain—control research opportunities and future directions.
        Ann Emerg Med. 1996; 27: 474-478
        • Van der Roost D
        Intervention en soins infirmiers. Face à la douleur aux urgences: action.
        Réan Urg. 1993; 2: 352-356
        • Berthier F
        • Le Conte P
        • Garrec F
        • Potel G
        • Baron D
        Analyse de la prise en charge de la douleur aiguë dans un service d’accueil et d’urgence.
        Réan Urg. 1998; 7: 281-285
        • Boureau F
        • Luu M
        • Héricord P
        Evaluation de l’intensité de la douleur au service d’accueil et des urgences, intér.
        Réan Urg. 1993; 2: 321-327
        • Hirsowski F
        • Boureau F
        Nécessité d’évaluer la douleur en pratique quotidienne.
        Douleurs. 2000; 1: 1
        • Ducharme J
        Emergency pain management.
        J Emerg Med. 1994; 12: 855-866
        • Yealy D.M
        Prehospital pain research.
        Ann Emerg Med. 1996; 27: 471-473
        • Berthier F
        • Potel G
        • Leconte P
        Comparative study of methods of measuring acute pain intensity in a ED.
        Am J Emerg Med. 1998; 16: 132-136
        • Huskisson E.C
        Measurement of pain.
        Lancet. 1974; 2: 1127-1131
        • Boureau F
        • Luu M
        • Gay C
        Les échelles d’évaluation de la symptomatologie douloureuse chronique.
        Med Hyg. 1982; 40: 3797-3805
        • Ricard-Hibon A
        • Leroy N
        • Magne M
        • Leberre A
        • Chollet C
        • Marty J
        Evaluation de la douleur aiguë en médecine préhospitalière.
        Ann Fr Anesth Réanim. 1997; 16: 945-949