- Email: [email protected]
“Feeling better” or “feeling well” in usual care of hip and knee osteoarthritis pain: Determination of cutoff points for patient acceptable symptom state (PASS) and minimal clinically important improvement (MCII) at rest and on movement in a national multicenter cohort study of 2414 patients with painful osteoarthritis
Ò
PAIN 154 (2013) 248–256
www.elsevier.com/locate/pain
‘‘Feeling better’’ or ‘‘feeling well’’ in usual care of hip and knee osteoarthritis pain: Determination of cutoff points for patient acceptable symptom state (PASS) and minimal clinically important improvement (MCII) at rest and on movement in a national multicenter cohort study of 2414 patients with painful osteoarthritis Serge Perrot a,⇑, Philippe Bertin b a b
Service de Médecine Interne et Thérapeutique, Hôtel Dieu, Assistance Publique Hôpitaux de Paris, Université Paris Descartes, INSERM U 987, Paris, France Service de Rhumatologie, CHU Dupuytren, Limoges, France
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
a r t i c l e
i n f o
Article history: Received 1 May 2012 Received in revised form 21 October 2012 Accepted 24 October 2012
Keywords: Acceptability Clinically relevance Osteoarthritis Pain
a b s t r a c t Patient-reported outcome measures are being developed for more relevant assessments of pain management. The patient acceptable symptom state (PASS) (‘‘feeling well’’) and the minimal clinically important improvement (MCII) (‘‘feeling better’’) have been determined in clinical trials, but not in daily pain management. We carried out a national multicenter cohort study of patients over the age of 50 years with painful knee osteoarthritis (KOA) or hip osteoarthritis (HOA) who had visited their general practitioner and required treatment for more than 7 days. Overall, 2414 patients (50.2% men, mean age 67.3 years, body mass index 27.9 kg/m2, 33.5% with HOA) were enrolled by 1116 general practitioners. After 7 days of treatment, PASS was estimated on a numerical rating scale as 4 at rest and 5 on movement, for both HOA and KOA, above the PASS threshold in clinical trials. In KOA, PASS was more frequently reached in men and younger people with less pain at rest and on movement, and in patients specifically seeking an improvement during sport activities. In HOA, PASS was most frequently reached in patients with low levels of pain at risk and in nonobese patients. MCII was 1 numerical rating scale point after 7 days of usual treatment. This improvement is smaller than that recorded in randomized controlled trials, and was the same for both sites, both at rest and on movement. In conclusion, patient-reported outcome values in daily practice differ from those in clinical trials, and their determinant factors may depend on the site of osteoarthritis. Assessments of the treatment of painful osteoarthritis should be adapted to the characteristics and daily life of the patient, to personalize patient management. Ó 2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
1. Introduction Interest in patient-reported outcomes (PROs) for the assessment of chronic conditions is increasing. Two complementary concepts reflecting a meaningful clinical response from the patient’s perspective have recently been developed in rheumatology and discussed at Outcome Measures in Rheumatology conferences (OMERACT 8 [23] and OMERACT 10 [18]). The relevance of these concepts in rheumatology has been confirmed through clinical trials [5,9]. The minimal clinically important improvement (MCII) ⇑ Corresponding author. Address: Service de Médecine Interne et Thérapeutique, Hôtel Dieu, 1 Place du Parvis Notre Dame, 7500 Paris, France. Tel.: +33 1 4234 8449; fax: +33 1 4234 8588. E-mail address: [email protected] (S. Perrot).
[11], defined as the smallest change in a patient’s symptom score given a substantial improvement in the patient’s condition, encompasses the concept of ‘‘feeling better’’ [3,21]. The patient acceptable symptom state (PASS), defined as the symptom score beyond which patients consider themselves well, reflects the concept of ‘‘feeling well’’ [5]. PASS assessment requires the clinician to ask the patient a single question requiring a yes-or-no answer: ‘‘Considering all the different ways in which your diseases affects you, if you were to remain in this state for the next few months, would you consider your current state to be satisfactory?’’ These 2 complementary concepts can be used to determine the proportion of patients considering their disease to be in an acceptable state (PASS) and the proportion of patients experiencing a major improvement in their condition (MCII) [12]. Both of these concepts have been shown to be important in rheumatology [23], but only a
0304-3959/$36.00 Ó 2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2012.10.017
Ò
249
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
few studies have focused on osteoarthritis (OA), all of them randomized controlled studies [5,21,22]. The cutoff points for PASS and MCII were determined on the basis of the results of clinical trials on selected patients with defined treatments, with a 0-to-100 mm visual analog scale (VAS). Furthermore, these cutoff points have been defined only for knee OA (KOA). The aim of our study was to determine the cutoff points for PASS and MCII in real life, and to compare these values between patients with painful KOA and hip OA (HOA), after 7 days of usual care (with or without drugs) delivered by general practitioners. We also compared PASS and MCII cutoff points in different situations (ie, at rest and on movement) and as a function of the patients’ priorities concerning their disability. This concept of disability-related patient priorities is derived from the McMaster-Toronto Arthritis Preference disability questionnaire [24]. It defines the activities that the patients cannot perform due to health status that they consider to be the most important. Five domains of activities were defined on the basis of the McMaster-Toronto Arthritis Preference disability questionnaire concept: household activities, sporting activities, outdoor activities, social activities, leisure activities. Finally, we compared the independent baseline demographic determinants of PASS and MCII for HOA and KOA. 2. Materials and methods We carried out a prospective, multicenter, observational cohort study in general practice in France between October 2009 and May 2010. The study was approved by the French Comité Consultatif sur le Traitement de l’Information en matière de Recherche dans le domaine de la Santé and the Commission Nationale Informatique et Liberté. Written informed consent was obtained from all patients before inclusion. The study design included a baseline visit, a 1-week treatment phase, and a final visit at day 7, the physicians being free to choose the nature of treatment prescribed for each patient. Consecutive outpatients age 50 years and over consulting their GP for painful HOA and/or KOA (both defined by the radiological and/or clinical criteria of the American College of Rheumatology [1]) with a pain intensity over the previous 24 hours of at least 3 of 10 on a numerical rating scale (NRS) of 0 to 10 were eligible for inclusion in the study. For recruitment, the patients had to require analgesic treatment for at least 7 days and to be capable of completing a patient diary. Patients treated with analgesics or nonsteroidal anti-inflammatory drugs in the 5 days before inclusion or who had received a joint injection at any time during the 4 weeks before inclusion could not be included. We also excluded patients suffering from any other concomitant painful disease likely to interfere with the assessment of PASS or for whom daily activity could not be assessed. At inclusion on day 0, the investigators collected demographic and disease data. Pain intensity at rest and on movement during the previous 24 hours was assessed on an NRS, and the primary disability, defined as the daily activity the patients most wanted to be improved (household, outdoor, sports, leisure, or social activities) was determined. Information on treatments was also collected. At inclusion, during the first visit, an analgesic treatment was started based on physicians’ usual care. Analgesics were classified on the World Health Organization analgesic ladder in the 3 different levels/steps. At day 7, pain at rest and on movement during the previous 24 hours was again assessed and the patients’ opinions of their state was assessed by recording their answer (yes or no) to the question ‘‘Taking into account the pain due to your arthritis in the last 24 hours, if you were to remain in this state for the next few years, would you consider your state acceptable?’’ PASS at rest
and on movement was determined by the 75th percentiles of the distributions of pain intensity in the subgroup of patients considering their condition to be acceptable at day 7. PASS at rest and on movement was determined, with a mean of these 2 values determining the overall PASS. The patients’ opinions of their pain relief were recorded as a response of improved, unchanged, or aggravated to the question ‘‘Taking into account the pain due to your arthritis in the last 24 hours and the pain you experienced initially, how has your pain changed?’’ Patients who said that their pain had improved were asked to rate the level of improvement on a 5-point Likert scale (very large improvement to no improvement at all). MCII at rest and on movement was determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement to be at least moderate. We calculated that a sample size of 3450 patients was required to achieve an absolute precision of 1.7% for the PASS threshold being reached by 30% of patients, assuming that PASS could be estimated for 80% of the patients. The proportions of patients with a PASS were determined with 95% confidence intervals (CI) for HOA and KOA. The factors independently accounting for PASS in patients with HOA and KOA were determined by stepwise logistic regression, with an entry level of 10% and a maintenance level of 15%, with selection of the explanatory variables from the set of variables significant at the 20% level when comparing the characteristics of patients with and without a PASS in v2 or Fisher exact tests for categorical data, Student t tests, or Wilcoxon nonparametric tests for quantitative data. Statistical analysis was performed with SAS version 9.1 for Windows (SAS Institute Inc., Cary, NC, USA). 3. Results 3.1. Eligible patients The study was proposed, by mailing, to a sample of 8669 French GPs, 1116 of whom (87.5% men, mean [SD] age 52.3 [±7.0] years) agreed to take part in the study (Table 1) and recruited 3329 patients between October 2009 and May 2010. In total, 915 case report forms (27.5%) were excluded from the analyses, mostly due to missing critical data (missing patient diary [n = 191], patient diary completed before day 5 [n = 210] or after day 10 [n = 424], missing questionnaire [n = 53]) (Fig. 1). 3.2. Patients at inclusion We analyzed the data for 2414 patients. Their mean age was 67.3 (±8.7) years, and men (50.2%) and women (49.8%) were equally represented. Most of the patients suffered from KOA (1606, 66.5%), the mean duration of which was 5.4 (±4.3) years, and 808 (33.5%) had been suffering from HOA for a mean of 4.9 (±4.0) years. Most patients were overweight (25 had a body mass index [BMI] 6 30 kg/m2, 50.9%) or obese (BMI P 30 kg/m2, 25.6%). All kinds of daily activities were affected by the disease in most patients, with outdoor (44.5%) and household (27.9%)
Table 1 Physician characteristics. (N = 1116) Age in years, mean (SD) Male Practice Large cities Small cities Country
52.3 (7.0) 87.5% 49.2% 26.1% 24.7%
Ò
250
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 2 Patient characteristics at baseline. Hip osteoarthritis (N = 808)
Knee osteoarthritis (N = 1606)
Total (N = 2414)
68.0 (8.2)
66.9 (9.0)
67.3 (8.7)
Age in years, mean (SD) Male Body mass index in kg/m2 (SD) <18.5 18.5–25 25–30 P30 Household Outdoor Sports Leisure Social activities
49.3% 27.3 (3.9)
50.7% 28.2 (4.5)
50.2% 27.9 (4.3)
0.3% 27.2% 53.1% 19.5% 63.8% 95.4% 87.9% 92.5% 59.5%
0.3% 21.2% 49.8% 28.7% 56.5% 94.6% 85.4% 89.7% 50.8%
0.3% 23.2% 50.9% 25.6% 59.0% 94.9% 86.2% 90.7% 53.7%
Most affected activities* Household Outdoor Sports Leisure Social activities
31.8% 42.3% 10.8% 10.3% 4.8%
26.0% 45.6% 11.7% 12.3% 4.5%
27.9% 44.5% 11.4% 11.6% 4.6%
43.6% 75.5% 27.5% 25.7% 2.4%
43.7% 72.5% 29.2% 28.6% 13.5%
43.6% 73.5% 28.6% 27.6% 9.7%
Treatments Level 1 analgesic Level 2 analgesic Other drugs Physiotherapy Intra-articular injections *
Assessed for 754 patients with hip osteoarthritis and 1488 patients with knee osteoarthritis.
Fig. 1. Patient flowchart: included patients.
activities being the activities the patients most wanted to improve. Most of the patients required level 2 analgesia (73.5%), but 43.6% used level 1 analgesics, 28.6% used other drugs, and 27.6% underwent physiotherapy. Mean (SD) pain intensity, assessed on an NRS of 0 to 10, was 5.1 (1.9) at rest and 7.0 (1.4) on movement (Table 2). 3.3. Change in pain after 7 days of treatment Most of the patients (89.6%) considered their pain at day 7 to have improved since baseline, the improvement being rated moderately large, large, or very large by 93.6% of these patients. Current
pain state was considered acceptable by 70.2% of the patients. Pain intensity had actually decreased in these patients and was rated 3.2 (1.8) at rest and 4.4 (1.9) on movement on an NRS from 0 to 10 (Table 3). The PASS, corresponding to the 75th percentile of the distribution of pain intensity on day 7 in the subgroup of patients who considered their state to be acceptable, was 4 points at rest (Fig. 2) and 5 on movement (Fig. 3). The PASS was reached for 79.4% of the patients at rest, 74.7% on movement, and 69.5% in both situations (9.9% only at rest and 5.2% only on movement). The proportion of patients reaching PASS both at rest and on movement was similar for HOA (68.3%) and KOA (70.1%). The MCII, corresponding to the 75th percentile of the distribution of pain intensity difference (day 0 to day 7) for patients considering their improvement in pain on day 7 to be at least moderately large, was 1 NRS point both at rest (Fig. 4) and on movement (Fig. 5). The MCII was reached by 80.0% of the patients for pain at rest, 86.3% for pain on movement, and 76.6% both at rest and on movement (3.4% only at rest and 9.7% only on movement). MCII and PASS were very strongly linked, mostly on movement: 97.7% of the patients reaching the PASS on movement also reported a MCII, and at rest, 87.8% of the patients reaching the PASS reported a MCII (P < .001). Finally, 89.2% of the patients reaching PASS both at rest and on movement reported a MCII 3.4. Independent baseline demographic and disability determinants of PASS The factors related to pain status being considered unacceptable (according to PASS cutoff points) in patients with HOA were high levels of pain at rest (odds ratio [95% CI]: 4.00 [1.75 to 9.09] for an NRS score of 5 to 7 and 6.25 [2.38 to 16.67] for an NRS score of 8 to 10 vs NRS score <3 and obesity (BMI P 30 kg/m2: 1.64 [1.02 to 2.63] vs [18.5 to 25] kg/m2) (Table 4).
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 3 Patient characteristics on day 7.
Pain at rest, mean numerical rating scale score (SD) Pain while moving, mean numerical rating scale score (SD) Acceptable condition according to the patient Condition improved, day 7 vs baseline Improvement Very large Large Moderately large Small Very small Compliance with the doctor’s instructions Patients reaching PASS at rest and on movement Patients reaching PASS at rest (PASS cutoff point at rest = 4) Patients reaching PASS on movement (PASS cutoff point on movement = 5) Patients reaching MCII at rest and on movement Patients reaching MCII at rest (MCII cutoff point at rest = 1) Patients reaching MCII on movement (MCII cutoff point on movement = 1)
Hip osteoarthritis (N = 808)
Knee osteoarthritis (N = 1606)
Total (N = 2414)
3.2 (1.8)
3.2 (1.9)
3.2 (1.8)
4.4 (1.8)
4.4 (1.9)
4.4 (1.9)
71.1%
69.7%
70.2%
90.7%
89.1%
89.6%
16.1% 42.3% 33.8% 7.8% – 98.2%
15.2% 44.7% 34.5% 5.6% – 97.4%
15.5% 43.9% 34.2% 6.3% 97.7%
68.3%
70.1%
69.5%
79.0%
79.6%
79.4%
74.8%
74.7%
74.7%
75.7%
77.0%
76.6%
78.7%
80.7%
80.0%
87.5%
85.7%
86.3%
MCII = minimal clinically important improvement; PASS = patient acceptable symptom state.
In patients with KOA, PASS was less frequently reached in female patients (OR [95% CI]: 1.45 [1.12 to 1.89]), in patients more than age 75 years (1.61 [1.11 to 2.27] vs [48 to 60]), with high levels of pain at rest (2.08 [1.20 to 3.57] for an NRS score of 5 to 7 and
251
4.17 [2.22 to 7.69] for an NRS score of 8 to 10 vs NRS score <3) and on movement (3.85 [1.75 to 8.33] for an NRS score of 8 to 10) vs NRS score 3 to 4, but PASS was most frequently reached when sport was the activity most affected by the disease (0.55 [0.35 to 0.86] vs outdoor activities) (Table 5, Fig. 6). 3.5. Independent baseline demographic and disability determinants of MCII In patients with HOA, the odds of reaching MCII were significantly higher (P < .001) for patients suffering from higher levels of pain at rest (OR [95% CI] vs NRS score <3: 15.49 [5.88 to 40.81] for an NRS score of 8 to 10, 8.18 [4.55 to 14.69] for an NRS score of 5 to 7, and 3.93 [2.22 to 6.97] for an NRS score of 3 to 4). Similarly, higher levels of pain on movement increased the odds of reaching MCII (OR [95%CI] vs an NRS score of 3 to 4: 3.21 [1.28 to 8.08] for an NRS score of 8 to 10, 3.73 [1.53 to 9.10] for an NRS score of 5 to 7). Obese patients (BMI P 30 kg/m2) were significantly less likely to achieve an MCII than patients with normal BMI (18.5 to 25) kg/m2) (OR [95% CI]: 0.48 [0.29 to 0.81]) (Table 6). Similar results were recorded for patients with KOA. Higher pain intensity at rest significantly increased (P < .001) the odds of reaching MCII (OR [95% CI] vs an NRS score <3: 4.43 [2.52 to 7.81] for an NRS score of 8 to 10, 5.08 [3.40 to 7.59] for an NRS score of 5 to 7, and 3.17 [2.11 to 4.75] for an NRS score of 3 to 4. High levels of pain on movement increased the chances of reaching MCII (OR [95% CI] vs an NRS score of 3 to 4: 2.38 [1.38 to 4.09] for an NRS score of 8 to 10 and 2.50 [1.51 to 4.13] for an NRS score of 5 to 7) (Table 7). 4. Discussion We report several observations relevant to the use of PROs in the daily management of chronic painful conditions. PASS has been defined as the value beyond which patients consider themselves to be well, and MCII has been defined as the minimal change beyond which patients consider themselves to be better. These concepts have been widely used in clinical trials for several conditions, including HOA and KOA [5,9,21,22], rheumatoid arthritis [8], ankylosing spondylitis [13,20], rotator cuff syndrome [19], and hip replacement [2,7]. This study is, to our knowledge, the first to
Fig. 2. PASS at rest determined by the 75th percentiles of the distributions of the pain intensity in the subgroup of patients rating their condition as Acceptable at day 7.
Ò
252
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
Fig. 3. PASS on movement were determined by the 75th percentiles of the distributions of the pain intensity in the subgroup of patients rating their condition as Acceptable at day 7.
MCII at rest: patients improved at rest have at least a variation of -1 between D0 and D7 100% 90% 80% 75%
% of patients
-2 -1 0 Change in pain intensity
1
Fig. 4. MCII at rest determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement as at least moderate.
investigate PASS and MCII for patients with painful lower limb OA in primary care settings rather than in a randomized clinical trial. In this prospective 7-day study, we applied the concepts of PASS and MCII to patients with painful HOA and KOA visiting their general practitioner, whatever the treatment and advice provided. We used the 75th percentile approach for analysis. This is the approach most widely used in studies of OA, the receiver-operator characteristics approach being used mostly for other conditions, such as rheumatoid arthritis [8] and ankylosing spondylitis [20]. The cutoff points we derived can be seen as threshold values for considering patients to be satisfied with their pain status or responding to the treatment prescribed by their general practitioners, whatever the treatment prescribed.
4.1. PASS cutoff points This report highlights the similarity of PASS cutoff points for HOA and KOA after medical management—4 at rest and 5 on movement, on an NRS—suggesting that KOA and HOA may be pooled for analysis in cohort studies. This finding is not consistent with the results for OA acceptability after joint replacement in a previous study, in which Escobar et al. demonstrated that the cutoff for PASS was lower 1 year after hip replacement than 1 year after knee replacement [7]. Our results also show that the PASS cutoff points for an observational cohort in daily practice (4 and 5 points for pain at rest and on movement, respectively, on an NRS from 0 to 10) are slightly higher than those reported in clinical trials: in selected pa-
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
253
Fig. 5. MCII on movement determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement as at least moderate.
Table 4 Independent factors accounting for a lack of acceptability of osteoarthritis pain (according to patient acceptable symptom state cutoff points; hip osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Age, years 48–60 60–65 65–70 70–75 P75
1 (ref) 0.88 0.81 1.31 1.51
0.50–1.56 0.46–1.41 0.78–2.22 0.90–2.56
.675 .449 .305 .114
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 2.38 4.00 6.25
1.00–5.55 1.75–9.09 2.38–16.67
.051 .001 <.0002
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 1.89 1.72
0.24–1.78 0.63–4.76
.403 .285
Body mass index (kg/m2) <18.5 18.5–25 25–30 P30
1.89 1 (ref) 0.91 1.64
0.11–33.33
.659
0.62–1.35 1.02–2.63
.636 .040
tients with KOA, PASS was estimated at 32 mm for pain intensity, 32 mm for patient’s global assessment of disease activity, and 31 mm for the Western Ontario and McMaster Universities questionnaire [5]. Several authors have demonstrated that receiveroperator characteristics analysis yields lower cutoff points for PASS [7,21] than the 75th percentile approach used here, in clinical trials on patients with OA [5]. In fact, in daily practice, NRS is more frequently used than VAS, even if less accurate, and it looks more clinically relevant to use a 4/10 NRS cutoff PASS point. When using a verbal scale, these cutoff points correspond to the border between mild and moderate, thus it may be considered that a treatment target should be no worse than mild pain. In our study, patients were included if they were experiencing pain at an intensity of more than 3 on an NRS from 0 to 10, corresponding to 30 mm on the VAS used in clinical trials. Our higher thresholds cannot therefore
be accounted for by a difference in the intensity of pain at inclusion. Our findings also show that the acceptability of OA pain is not homogeneous, with different levels of acceptability recorded for pain at rest and pain on movement. This aspect should be taken into account, as pain may differ considerably at rest and on movement, particularly when pain fluctuations are considered, and this may have an impact on analyses [10]. 4.2. Independent baseline determinants of PASS cutoff points Despite the similarity of the PASS thresholds determined here for KOA and HOA, these thresholds had different independent demographic and patient priority baseline determinants. We identified 5 independent baseline determinants for KOA PASS thresholds: age, sex, level of pain at rest, level of pain on movement, and the disability-related priority of the patient. By contrast, we identified only 2 independent baseline determinants for HOA PASS thresholds: level of pain at rest and BMI. Pain intensity levels at rest and on movement influenced the PASS thresholds for KOA, whereas only pain intensity at rest influenced the PASS threshold for HOA. Patient disability-related priority was an independent determinant factor only for the KOA PASS threshold. This suggests that even if pain levels do not differ significantly between HOA and KOA, the acceptability of HOA pain may be influenced by a larger number of independent factors than that of KOA pain. This finding differs from that for joint replacement acceptability, for which sex, age, and BMI do not seem to influence PASS cutoff points [7]. It also conflicts with the data obtained in randomized controlled studies of HOA and KOA, in which age, sex, and disease duration were found to have no effect on the frequency of PASS [21]. This may reflect the different coping strategies used by patients with HOA and KOA in real life [15]. It is also consistent with the findings recently reported by Escobar et al. [7], showing that patients undergoing hip replacement displayed a greater improvement in pain and functionality 1 year after transplantation, with a faster functional improvement, than patients undergoing knee replacement. In other studies, on ankylosing spondylitis for example [13], the duration of symptoms directly affects the PASS threshold. By contrast, in our study, OA duration did not influence the PASS threshold.
Ò
254
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
Table 5 Independent factors accounting for a lack of acceptability of osteoarthritis pain (according to patient acceptable symptom state cutoff point; knee osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Sex Male Female
1 (ref) 1.45
1.12–1.89
.004
Age, years <48–60 60–65 65–70 70–75 P75
1 (ref) 0.91 1.08 1.19 1.61
0.62–1.33 0.73–1.59 0.81–1.72 1.11–2.27
.617 .721 .369 .010
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 1.35 2.08 4.17
0.76–2.38 3.57–1.20 2.22–7.69
.314 .001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 1.52 3.85
0.69–3.33 1.75–8.33
.296 <.001
Most affected activities Household Outdoor Sports Leisure Social activities
1.09 1 (ref) 0.55 0.79 1.27
0.81–1.47
.578
0.35–0.86 0.53–1.19 0.73–2.22
.001 .268 .395
1 (ref) 0.68
0.46–1.02
.061
0.78
0.49–1.25
.298
0.97
0.66–1.43
.898
Disease-modifying treatment Level 1 analgesics Level 2 analgesics, with or without level 1 analgesics Level 1 or 2 analgesic with another drug Drugs plus nonpharmacological therapies
This may be because most patients had had symptoms for a long time and because it was difficult to define the exact time of onset of OA disease.
4.3. MCII cutoff points Our results for chronic OA pain show that MCII levels are similar for HOA and KOA, at rest and on movement, with an estimated 1-point decrease in the score attributed on an NRS running from 0 to 10. These findings are similar to those for other studies [6,16], but the improvement is smaller than that reported by Tubach et al. [22]: MCII of 19.9 mm for KOA and 15.3 mm for HOA on a VAS (0 to 100). This difference may reflect differences in the clinical setting: Pope et al. [16] demonstrated that MCII was lower in clinical practice than in randomized clinical trials. The likelihood of reaching the MCII was correlated with pain intensity, both at rest and on movement: the higher the pain intensity at baseline, the greater the proportion of patients reaching the MCII. These findings are consistent with those of other studies on OA [14,20] and arthritis [17] that have demonstrated that MCII depends on baseline pain score. In our study, age, sex, and the disability-related priority of the patients had no effect on MCII, whereas BMI was an independent determining factor for KOA, with obese patients less frequently achieving the MCII than patients with a normal BMI. These findings are consistent with those of other studies on MCII in functional index, showing that MCII is context-specific and not a fixed attribute [25]. 4.4. Limitations of the study This was an observational study, in general practice, with a short (7 days) follow-up period. The acceptability of OA pain may be different in chronic care and for longer durations, due to the higher frequency of adverse events or the larger proportion of patients lost to follow-up, with longer follow-up periods. For reasons of feasibility, only limited data were collected at the start of the study, principally demographic data, pain intensities, and the patients’ disability-related priorities. The analysis of PASS and MCII determinants was therefore limited to this restricted set of variables. We did not assess the patients’ psychological determinants, their expectations, or the extent to which they agreed with their physician on the management of the disease. It would be useful
Fig. 6. Proportions of patients reaching PASS at rest, on movement and in both situations, as a function of their disability-related priorities.
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 6 Independent factors accounting for improvement (according to the minimal clinically important improvement cutoff; hip osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 3.93 8.18 15.49
2.22–6.97 4.55–14.69 5.88–40.81
<.001 <.001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 3.73 3.21
1.53–9.10 1.28–8.08
.004 .013
Body mass index, kg/m2 <18.5 18.5–25 25–30 P30
0.22 1 (ref) 0.87 0.48
0.01–3.84
.299
0.57–1.34 0.29–0.81
.527 .006
Table 7 Independent factors accounting for improvement (according to minimal clinically important improvement cutoff) (knee osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 3.17 5.08 4.43
2.11–4.75 3.40–7.59 2.52–7.81
<.001 <.001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 2.50 2.38
1.51–4.13 1.38–4.09
.0004 .002
Body mass index, kg/m2 <18.5 18.5–25 25–30 P30
0.07 1 (ref) 1.06 0.80
0.01–0.76
.024
0.77–1.46 0.57–1.13
.718 .215
to collect such additional information in future studies to improve our understanding of the acceptability of chronic painful conditions such as OA because pain severity in OA has been shown to be associated with both demographic and psychological variables [4]. 4.5. Conclusions In primary care settings, patients with lower limb OA may be considered to be in an acceptable symptomatic state if they assign a score of <4 to their pain on an NRS from 0 to 10 (40 mm on a VAS) at rest and a score of <5 (50 mm on a VAS) on movement, corresponding to no worse than mild pain. The independent determinants of PASS in lower limb OA differ between HOA and KOA. For patients with KOA, the cutoff point for PASS is influenced by age, sex, and pain intensity at rest and on movement, whereas for patients with HOA, it is determined BMI and pain at rest. After pain management in daily practice, the cutoff for the MCII was a decrease of 1 point on an NRS from 0 to 10 ( 10 mm on a VAS). The odds of reaching MCII increased with pain intensities at rest and movement for both KOA and HOA and were also influenced by BMI in KOA. Thus, the PASS and MCII concepts are relevant to daily practice. They should be used in daily clinical practice, and not only in clinical trials, for a more personalized management of chronic pain states. PASS and MCII could be used in the future for the routine assessment of PROs for analgesic treatment.
255
Acknowledgments This study was developed with an unrestricted grant from SANOFI France. This article was prepared with the help of a medical writer, Dr. Christian Kempf, from CEGEDIM and Julie Sappa for English corrections. References [1] Altman RD, Asch D, Bloch G, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–49. [2] Arden NK, Kiran A, Judge A, Biant LC, Javaid MK, Murray DW, et al. What is a good patient reported outcome after total hip replacement? Osteoarthritis Cartilage 2011;19:155–62. [3] Beaton DE, Boers M, Wells GA. Many faces of the minimal clinically important difference (MCID): a literature review and directions for future research. Curr Opin Rheumatol 2002;14:109–14. [4] Creamer P, Lethbridge-Cejku M, Hochberg M. Determinants of pain severity in knee osteoarthritis: effect of demographic and psychosocial variables using 3 pain measures. J Rheumatol 1999;26:1785–92. [5] Dougados M, Moore A, Yu S, Gitton X. Evaluation of the patient acceptable symptom state in a pooled analysis of two multicentre, randomised, doubleblind, placebo-controlled studies evaluating lumiracoxib and celecoxib in patients with osteoarthritis. Arthritis Res Ther 2007;9:R11. [6] Ehrich EW, Davies GM, Watson DJ, Bolognese JA, Seidenberg BC, Bellamy N. Minimal perceptible clinical improvement with the Western Ontario and McMaster Universities osteoarthritis index questionnaire and global assessments in patients with osteoarthritis. J Rheumatol 2000;27:2635–41. [7] Escobar A, Gonzalez M, Quintana JM, Vrotsou K, Bilbao A, Herrera-Espiñeira C, Garcia-Perez L, Aizpuru F, Sarasqueta C. Patient acceptable symptom state and OMERACT-OARSI set of responder criteria in joint replacement. Identification of cut-off values. Osteoarthritis Cartilage 2011;20:87–92. [8] Heiberg T, Kvien TK, Mowinckel P, et al. Identification of disease activity and health status cut-off points for the symptom state acceptable to patients with rheumatoid arthritis. Ann Rheum Dis 2008;67:967–71. [9] Hochberg MC, Wohlreich M, Gaynor P, Hanna S, Risser R. Clinically relevant outcomes based on analysis of pooled data from 2 trials of duloxetine in patients with knee osteoarthritis. J Rheumatol 2012;39:352–8. [10] Hutchings A, Calloway M, Choy E, Hooper M, Hunter DJ, Jordan JM, Zhang Y, Baser O, Long S, Palmer L. The Longitudinal Examination of Arthritis Pain (LEAP) study: relationships between weekly fluctuations in patient-rated joint pain and other health outcomes. J Rheumatol 2007;34:2291–300. [11] Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989;10:407–15. [12] Kvien TK, Heiberg T, Hagen KB. Minimal clinically important improvement/ difference (MCII/MCID) and patient acceptable symptom state (PASS): what do these concepts mean? Ann Rheum Dis 2007;66:40–1. [13] Maksymowych WP, Gooch K, Dougados M, Wong RL, Chen N, Kupper H, van der Heijde D. Thresholds of patient-reported outcomes that define the patient acceptable symptom state in ankylosing spondylitis vary over time and by treatment and patient characteristics. Arthritis Care Res 2010;62:826–34. [14] Ornetti P, Dougados M, Paternotte S, Logeart I, Gossec L. Validation of a numerical rating scale to assess functional impairment in hip and knee osteoarthritis: comparison with the WOMAC function scale. Ann Rheum Dis 2011;70:740–6. [15] Perrot S, Poiraudeau S, Kabir M, Bertin P, Sichere P, Serrie A, Rannou F. Active or passive pain coping strategies in hip and knee osteoarthritis? Results of a national survey of 4719 patients in a primary care setting. Arthritis Rheum 2008;59:1555–62. [16] Pope JE, Khanna D, Norrie D, Ouimet JM. The minimally important difference for the health assessment questionnaire in rheumatoid arthritis clinical practice is smaller than in randomized controlled trials. J Rheumatol 2009;36:254–9. [17] Stauffer ME, Taylor SD, Watson DJ, Peloso PM, Morrison A. Definition of nonresponse to analgesic treatment of arthritic pain: an analytical literature review of the smallest detectable difference, the minimal detectable change, and the minimal clinically important difference on the pain visual analog scale. Int J Inflamm 2011;23:19–26. [18] Strand V, Boers M, Idzerda L, Kirwan JR, Kvien TK, Tugwell PS, Dougados M. It’s good to feel better but it’s better to feel good and even better to feel good as soon as possible for as long as possible. Response criteria and the importance of change at OMERACT 10. J Rheumatol 2011;38:1720–7. [19] Tashjian RZ, Deloach J, Porucznik CA, Powell AP. Minimal clinically important differences (MCID) and patient acceptable symptomatic state (PASS) for visual analog scales (VAS) measuring pain in patients treated for rotator cuff disease. J Shoulder Elbow Surg 2009;18:927–32. [20] Tubach F, Pham T, Skomsvoll JF, et al. Stability of the patient acceptable symptomatic state over time in outcome criteria in ankylosing spondylitis. Arthritis Rheum 2006;55:960–3. [21] Tubach F, Ravaud P, Baron G, Falissard B, Logeart I, Bellamy N, et al. Evaluation of clinically relevant states in patient-reported outcomes in knee and hip
256
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
osteoarthritis: the patient acceptable symptom state. Ann Rheum Dis 2005;64:34–7. [22] Tubach F, Ravaud P, Baron G, Falissard B, Logeart I, Bellamy N, et al. Evaluation of clinically relevant changes in patient-reported outcomes in knee and hip osteoarthritis: the minimal clinically important improvement. Ann Rheum Dis 2005;64:29–33. [23] Tubach F, Ravaud P, Beaton D, Boers M, Bombardier C, Felson DT, et al. Minimal clinically important improvement and patient acceptable symptom state for
subjective outcome measures in rheumatic disorders. J Rheumatol 2007;34:1188–93. [24] Tugwell P, Bombardier C, Buchanan WW, Goldsmith CH, Grace E, et al. The MACTAR Patient Preference Disability Questionnaire—an individualized functional priority approach for assessing improvement in physical disability in clinical trials in rheumatoid arthritis. J Rheumatol 1987;14:446–4451. [25] Wang YC, Hart DL, Stratford PW, Mioduski JE. Baseline dependency of minimal clinically important improvement. Phys Ther 2011;91:675–88.
PAIN 154 (2013) 248–256
www.elsevier.com/locate/pain
‘‘Feeling better’’ or ‘‘feeling well’’ in usual care of hip and knee osteoarthritis pain: Determination of cutoff points for patient acceptable symptom state (PASS) and minimal clinically important improvement (MCII) at rest and on movement in a national multicenter cohort study of 2414 patients with painful osteoarthritis Serge Perrot a,⇑, Philippe Bertin b a b
Service de Médecine Interne et Thérapeutique, Hôtel Dieu, Assistance Publique Hôpitaux de Paris, Université Paris Descartes, INSERM U 987, Paris, France Service de Rhumatologie, CHU Dupuytren, Limoges, France
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
a r t i c l e
i n f o
Article history: Received 1 May 2012 Received in revised form 21 October 2012 Accepted 24 October 2012
Keywords: Acceptability Clinically relevance Osteoarthritis Pain
a b s t r a c t Patient-reported outcome measures are being developed for more relevant assessments of pain management. The patient acceptable symptom state (PASS) (‘‘feeling well’’) and the minimal clinically important improvement (MCII) (‘‘feeling better’’) have been determined in clinical trials, but not in daily pain management. We carried out a national multicenter cohort study of patients over the age of 50 years with painful knee osteoarthritis (KOA) or hip osteoarthritis (HOA) who had visited their general practitioner and required treatment for more than 7 days. Overall, 2414 patients (50.2% men, mean age 67.3 years, body mass index 27.9 kg/m2, 33.5% with HOA) were enrolled by 1116 general practitioners. After 7 days of treatment, PASS was estimated on a numerical rating scale as 4 at rest and 5 on movement, for both HOA and KOA, above the PASS threshold in clinical trials. In KOA, PASS was more frequently reached in men and younger people with less pain at rest and on movement, and in patients specifically seeking an improvement during sport activities. In HOA, PASS was most frequently reached in patients with low levels of pain at risk and in nonobese patients. MCII was 1 numerical rating scale point after 7 days of usual treatment. This improvement is smaller than that recorded in randomized controlled trials, and was the same for both sites, both at rest and on movement. In conclusion, patient-reported outcome values in daily practice differ from those in clinical trials, and their determinant factors may depend on the site of osteoarthritis. Assessments of the treatment of painful osteoarthritis should be adapted to the characteristics and daily life of the patient, to personalize patient management. Ó 2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
1. Introduction Interest in patient-reported outcomes (PROs) for the assessment of chronic conditions is increasing. Two complementary concepts reflecting a meaningful clinical response from the patient’s perspective have recently been developed in rheumatology and discussed at Outcome Measures in Rheumatology conferences (OMERACT 8 [23] and OMERACT 10 [18]). The relevance of these concepts in rheumatology has been confirmed through clinical trials [5,9]. The minimal clinically important improvement (MCII) ⇑ Corresponding author. Address: Service de Médecine Interne et Thérapeutique, Hôtel Dieu, 1 Place du Parvis Notre Dame, 7500 Paris, France. Tel.: +33 1 4234 8449; fax: +33 1 4234 8588. E-mail address: [email protected] (S. Perrot).
[11], defined as the smallest change in a patient’s symptom score given a substantial improvement in the patient’s condition, encompasses the concept of ‘‘feeling better’’ [3,21]. The patient acceptable symptom state (PASS), defined as the symptom score beyond which patients consider themselves well, reflects the concept of ‘‘feeling well’’ [5]. PASS assessment requires the clinician to ask the patient a single question requiring a yes-or-no answer: ‘‘Considering all the different ways in which your diseases affects you, if you were to remain in this state for the next few months, would you consider your current state to be satisfactory?’’ These 2 complementary concepts can be used to determine the proportion of patients considering their disease to be in an acceptable state (PASS) and the proportion of patients experiencing a major improvement in their condition (MCII) [12]. Both of these concepts have been shown to be important in rheumatology [23], but only a
0304-3959/$36.00 Ó 2012 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.pain.2012.10.017
Ò
249
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
few studies have focused on osteoarthritis (OA), all of them randomized controlled studies [5,21,22]. The cutoff points for PASS and MCII were determined on the basis of the results of clinical trials on selected patients with defined treatments, with a 0-to-100 mm visual analog scale (VAS). Furthermore, these cutoff points have been defined only for knee OA (KOA). The aim of our study was to determine the cutoff points for PASS and MCII in real life, and to compare these values between patients with painful KOA and hip OA (HOA), after 7 days of usual care (with or without drugs) delivered by general practitioners. We also compared PASS and MCII cutoff points in different situations (ie, at rest and on movement) and as a function of the patients’ priorities concerning their disability. This concept of disability-related patient priorities is derived from the McMaster-Toronto Arthritis Preference disability questionnaire [24]. It defines the activities that the patients cannot perform due to health status that they consider to be the most important. Five domains of activities were defined on the basis of the McMaster-Toronto Arthritis Preference disability questionnaire concept: household activities, sporting activities, outdoor activities, social activities, leisure activities. Finally, we compared the independent baseline demographic determinants of PASS and MCII for HOA and KOA. 2. Materials and methods We carried out a prospective, multicenter, observational cohort study in general practice in France between October 2009 and May 2010. The study was approved by the French Comité Consultatif sur le Traitement de l’Information en matière de Recherche dans le domaine de la Santé and the Commission Nationale Informatique et Liberté. Written informed consent was obtained from all patients before inclusion. The study design included a baseline visit, a 1-week treatment phase, and a final visit at day 7, the physicians being free to choose the nature of treatment prescribed for each patient. Consecutive outpatients age 50 years and over consulting their GP for painful HOA and/or KOA (both defined by the radiological and/or clinical criteria of the American College of Rheumatology [1]) with a pain intensity over the previous 24 hours of at least 3 of 10 on a numerical rating scale (NRS) of 0 to 10 were eligible for inclusion in the study. For recruitment, the patients had to require analgesic treatment for at least 7 days and to be capable of completing a patient diary. Patients treated with analgesics or nonsteroidal anti-inflammatory drugs in the 5 days before inclusion or who had received a joint injection at any time during the 4 weeks before inclusion could not be included. We also excluded patients suffering from any other concomitant painful disease likely to interfere with the assessment of PASS or for whom daily activity could not be assessed. At inclusion on day 0, the investigators collected demographic and disease data. Pain intensity at rest and on movement during the previous 24 hours was assessed on an NRS, and the primary disability, defined as the daily activity the patients most wanted to be improved (household, outdoor, sports, leisure, or social activities) was determined. Information on treatments was also collected. At inclusion, during the first visit, an analgesic treatment was started based on physicians’ usual care. Analgesics were classified on the World Health Organization analgesic ladder in the 3 different levels/steps. At day 7, pain at rest and on movement during the previous 24 hours was again assessed and the patients’ opinions of their state was assessed by recording their answer (yes or no) to the question ‘‘Taking into account the pain due to your arthritis in the last 24 hours, if you were to remain in this state for the next few years, would you consider your state acceptable?’’ PASS at rest
and on movement was determined by the 75th percentiles of the distributions of pain intensity in the subgroup of patients considering their condition to be acceptable at day 7. PASS at rest and on movement was determined, with a mean of these 2 values determining the overall PASS. The patients’ opinions of their pain relief were recorded as a response of improved, unchanged, or aggravated to the question ‘‘Taking into account the pain due to your arthritis in the last 24 hours and the pain you experienced initially, how has your pain changed?’’ Patients who said that their pain had improved were asked to rate the level of improvement on a 5-point Likert scale (very large improvement to no improvement at all). MCII at rest and on movement was determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement to be at least moderate. We calculated that a sample size of 3450 patients was required to achieve an absolute precision of 1.7% for the PASS threshold being reached by 30% of patients, assuming that PASS could be estimated for 80% of the patients. The proportions of patients with a PASS were determined with 95% confidence intervals (CI) for HOA and KOA. The factors independently accounting for PASS in patients with HOA and KOA were determined by stepwise logistic regression, with an entry level of 10% and a maintenance level of 15%, with selection of the explanatory variables from the set of variables significant at the 20% level when comparing the characteristics of patients with and without a PASS in v2 or Fisher exact tests for categorical data, Student t tests, or Wilcoxon nonparametric tests for quantitative data. Statistical analysis was performed with SAS version 9.1 for Windows (SAS Institute Inc., Cary, NC, USA). 3. Results 3.1. Eligible patients The study was proposed, by mailing, to a sample of 8669 French GPs, 1116 of whom (87.5% men, mean [SD] age 52.3 [±7.0] years) agreed to take part in the study (Table 1) and recruited 3329 patients between October 2009 and May 2010. In total, 915 case report forms (27.5%) were excluded from the analyses, mostly due to missing critical data (missing patient diary [n = 191], patient diary completed before day 5 [n = 210] or after day 10 [n = 424], missing questionnaire [n = 53]) (Fig. 1). 3.2. Patients at inclusion We analyzed the data for 2414 patients. Their mean age was 67.3 (±8.7) years, and men (50.2%) and women (49.8%) were equally represented. Most of the patients suffered from KOA (1606, 66.5%), the mean duration of which was 5.4 (±4.3) years, and 808 (33.5%) had been suffering from HOA for a mean of 4.9 (±4.0) years. Most patients were overweight (25 had a body mass index [BMI] 6 30 kg/m2, 50.9%) or obese (BMI P 30 kg/m2, 25.6%). All kinds of daily activities were affected by the disease in most patients, with outdoor (44.5%) and household (27.9%)
Table 1 Physician characteristics. (N = 1116) Age in years, mean (SD) Male Practice Large cities Small cities Country
52.3 (7.0) 87.5% 49.2% 26.1% 24.7%
Ò
250
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 2 Patient characteristics at baseline. Hip osteoarthritis (N = 808)
Knee osteoarthritis (N = 1606)
Total (N = 2414)
68.0 (8.2)
66.9 (9.0)
67.3 (8.7)
Age in years, mean (SD) Male Body mass index in kg/m2 (SD) <18.5 18.5–25 25–30 P30 Household Outdoor Sports Leisure Social activities
49.3% 27.3 (3.9)
50.7% 28.2 (4.5)
50.2% 27.9 (4.3)
0.3% 27.2% 53.1% 19.5% 63.8% 95.4% 87.9% 92.5% 59.5%
0.3% 21.2% 49.8% 28.7% 56.5% 94.6% 85.4% 89.7% 50.8%
0.3% 23.2% 50.9% 25.6% 59.0% 94.9% 86.2% 90.7% 53.7%
Most affected activities* Household Outdoor Sports Leisure Social activities
31.8% 42.3% 10.8% 10.3% 4.8%
26.0% 45.6% 11.7% 12.3% 4.5%
27.9% 44.5% 11.4% 11.6% 4.6%
43.6% 75.5% 27.5% 25.7% 2.4%
43.7% 72.5% 29.2% 28.6% 13.5%
43.6% 73.5% 28.6% 27.6% 9.7%
Treatments Level 1 analgesic Level 2 analgesic Other drugs Physiotherapy Intra-articular injections *
Assessed for 754 patients with hip osteoarthritis and 1488 patients with knee osteoarthritis.
Fig. 1. Patient flowchart: included patients.
activities being the activities the patients most wanted to improve. Most of the patients required level 2 analgesia (73.5%), but 43.6% used level 1 analgesics, 28.6% used other drugs, and 27.6% underwent physiotherapy. Mean (SD) pain intensity, assessed on an NRS of 0 to 10, was 5.1 (1.9) at rest and 7.0 (1.4) on movement (Table 2). 3.3. Change in pain after 7 days of treatment Most of the patients (89.6%) considered their pain at day 7 to have improved since baseline, the improvement being rated moderately large, large, or very large by 93.6% of these patients. Current
pain state was considered acceptable by 70.2% of the patients. Pain intensity had actually decreased in these patients and was rated 3.2 (1.8) at rest and 4.4 (1.9) on movement on an NRS from 0 to 10 (Table 3). The PASS, corresponding to the 75th percentile of the distribution of pain intensity on day 7 in the subgroup of patients who considered their state to be acceptable, was 4 points at rest (Fig. 2) and 5 on movement (Fig. 3). The PASS was reached for 79.4% of the patients at rest, 74.7% on movement, and 69.5% in both situations (9.9% only at rest and 5.2% only on movement). The proportion of patients reaching PASS both at rest and on movement was similar for HOA (68.3%) and KOA (70.1%). The MCII, corresponding to the 75th percentile of the distribution of pain intensity difference (day 0 to day 7) for patients considering their improvement in pain on day 7 to be at least moderately large, was 1 NRS point both at rest (Fig. 4) and on movement (Fig. 5). The MCII was reached by 80.0% of the patients for pain at rest, 86.3% for pain on movement, and 76.6% both at rest and on movement (3.4% only at rest and 9.7% only on movement). MCII and PASS were very strongly linked, mostly on movement: 97.7% of the patients reaching the PASS on movement also reported a MCII, and at rest, 87.8% of the patients reaching the PASS reported a MCII (P < .001). Finally, 89.2% of the patients reaching PASS both at rest and on movement reported a MCII 3.4. Independent baseline demographic and disability determinants of PASS The factors related to pain status being considered unacceptable (according to PASS cutoff points) in patients with HOA were high levels of pain at rest (odds ratio [95% CI]: 4.00 [1.75 to 9.09] for an NRS score of 5 to 7 and 6.25 [2.38 to 16.67] for an NRS score of 8 to 10 vs NRS score <3 and obesity (BMI P 30 kg/m2: 1.64 [1.02 to 2.63] vs [18.5 to 25] kg/m2) (Table 4).
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 3 Patient characteristics on day 7.
Pain at rest, mean numerical rating scale score (SD) Pain while moving, mean numerical rating scale score (SD) Acceptable condition according to the patient Condition improved, day 7 vs baseline Improvement Very large Large Moderately large Small Very small Compliance with the doctor’s instructions Patients reaching PASS at rest and on movement Patients reaching PASS at rest (PASS cutoff point at rest = 4) Patients reaching PASS on movement (PASS cutoff point on movement = 5) Patients reaching MCII at rest and on movement Patients reaching MCII at rest (MCII cutoff point at rest = 1) Patients reaching MCII on movement (MCII cutoff point on movement = 1)
Hip osteoarthritis (N = 808)
Knee osteoarthritis (N = 1606)
Total (N = 2414)
3.2 (1.8)
3.2 (1.9)
3.2 (1.8)
4.4 (1.8)
4.4 (1.9)
4.4 (1.9)
71.1%
69.7%
70.2%
90.7%
89.1%
89.6%
16.1% 42.3% 33.8% 7.8% – 98.2%
15.2% 44.7% 34.5% 5.6% – 97.4%
15.5% 43.9% 34.2% 6.3% 97.7%
68.3%
70.1%
69.5%
79.0%
79.6%
79.4%
74.8%
74.7%
74.7%
75.7%
77.0%
76.6%
78.7%
80.7%
80.0%
87.5%
85.7%
86.3%
MCII = minimal clinically important improvement; PASS = patient acceptable symptom state.
In patients with KOA, PASS was less frequently reached in female patients (OR [95% CI]: 1.45 [1.12 to 1.89]), in patients more than age 75 years (1.61 [1.11 to 2.27] vs [48 to 60]), with high levels of pain at rest (2.08 [1.20 to 3.57] for an NRS score of 5 to 7 and
251
4.17 [2.22 to 7.69] for an NRS score of 8 to 10 vs NRS score <3) and on movement (3.85 [1.75 to 8.33] for an NRS score of 8 to 10) vs NRS score 3 to 4, but PASS was most frequently reached when sport was the activity most affected by the disease (0.55 [0.35 to 0.86] vs outdoor activities) (Table 5, Fig. 6). 3.5. Independent baseline demographic and disability determinants of MCII In patients with HOA, the odds of reaching MCII were significantly higher (P < .001) for patients suffering from higher levels of pain at rest (OR [95% CI] vs NRS score <3: 15.49 [5.88 to 40.81] for an NRS score of 8 to 10, 8.18 [4.55 to 14.69] for an NRS score of 5 to 7, and 3.93 [2.22 to 6.97] for an NRS score of 3 to 4). Similarly, higher levels of pain on movement increased the odds of reaching MCII (OR [95%CI] vs an NRS score of 3 to 4: 3.21 [1.28 to 8.08] for an NRS score of 8 to 10, 3.73 [1.53 to 9.10] for an NRS score of 5 to 7). Obese patients (BMI P 30 kg/m2) were significantly less likely to achieve an MCII than patients with normal BMI (18.5 to 25) kg/m2) (OR [95% CI]: 0.48 [0.29 to 0.81]) (Table 6). Similar results were recorded for patients with KOA. Higher pain intensity at rest significantly increased (P < .001) the odds of reaching MCII (OR [95% CI] vs an NRS score <3: 4.43 [2.52 to 7.81] for an NRS score of 8 to 10, 5.08 [3.40 to 7.59] for an NRS score of 5 to 7, and 3.17 [2.11 to 4.75] for an NRS score of 3 to 4. High levels of pain on movement increased the chances of reaching MCII (OR [95% CI] vs an NRS score of 3 to 4: 2.38 [1.38 to 4.09] for an NRS score of 8 to 10 and 2.50 [1.51 to 4.13] for an NRS score of 5 to 7) (Table 7). 4. Discussion We report several observations relevant to the use of PROs in the daily management of chronic painful conditions. PASS has been defined as the value beyond which patients consider themselves to be well, and MCII has been defined as the minimal change beyond which patients consider themselves to be better. These concepts have been widely used in clinical trials for several conditions, including HOA and KOA [5,9,21,22], rheumatoid arthritis [8], ankylosing spondylitis [13,20], rotator cuff syndrome [19], and hip replacement [2,7]. This study is, to our knowledge, the first to
Fig. 2. PASS at rest determined by the 75th percentiles of the distributions of the pain intensity in the subgroup of patients rating their condition as Acceptable at day 7.
Ò
252
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
Fig. 3. PASS on movement were determined by the 75th percentiles of the distributions of the pain intensity in the subgroup of patients rating their condition as Acceptable at day 7.
MCII at rest: patients improved at rest have at least a variation of -1 between D0 and D7 100% 90% 80% 75%
% of patients
-2 -1 0 Change in pain intensity
1
Fig. 4. MCII at rest determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement as at least moderate.
investigate PASS and MCII for patients with painful lower limb OA in primary care settings rather than in a randomized clinical trial. In this prospective 7-day study, we applied the concepts of PASS and MCII to patients with painful HOA and KOA visiting their general practitioner, whatever the treatment and advice provided. We used the 75th percentile approach for analysis. This is the approach most widely used in studies of OA, the receiver-operator characteristics approach being used mostly for other conditions, such as rheumatoid arthritis [8] and ankylosing spondylitis [20]. The cutoff points we derived can be seen as threshold values for considering patients to be satisfied with their pain status or responding to the treatment prescribed by their general practitioners, whatever the treatment prescribed.
4.1. PASS cutoff points This report highlights the similarity of PASS cutoff points for HOA and KOA after medical management—4 at rest and 5 on movement, on an NRS—suggesting that KOA and HOA may be pooled for analysis in cohort studies. This finding is not consistent with the results for OA acceptability after joint replacement in a previous study, in which Escobar et al. demonstrated that the cutoff for PASS was lower 1 year after hip replacement than 1 year after knee replacement [7]. Our results also show that the PASS cutoff points for an observational cohort in daily practice (4 and 5 points for pain at rest and on movement, respectively, on an NRS from 0 to 10) are slightly higher than those reported in clinical trials: in selected pa-
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
253
Fig. 5. MCII on movement determined by the 75th percentile of the distribution of the pain intensity difference (day 0 to day 7) for patients considering their improvement as at least moderate.
Table 4 Independent factors accounting for a lack of acceptability of osteoarthritis pain (according to patient acceptable symptom state cutoff points; hip osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Age, years 48–60 60–65 65–70 70–75 P75
1 (ref) 0.88 0.81 1.31 1.51
0.50–1.56 0.46–1.41 0.78–2.22 0.90–2.56
.675 .449 .305 .114
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 2.38 4.00 6.25
1.00–5.55 1.75–9.09 2.38–16.67
.051 .001 <.0002
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 1.89 1.72
0.24–1.78 0.63–4.76
.403 .285
Body mass index (kg/m2) <18.5 18.5–25 25–30 P30
1.89 1 (ref) 0.91 1.64
0.11–33.33
.659
0.62–1.35 1.02–2.63
.636 .040
tients with KOA, PASS was estimated at 32 mm for pain intensity, 32 mm for patient’s global assessment of disease activity, and 31 mm for the Western Ontario and McMaster Universities questionnaire [5]. Several authors have demonstrated that receiveroperator characteristics analysis yields lower cutoff points for PASS [7,21] than the 75th percentile approach used here, in clinical trials on patients with OA [5]. In fact, in daily practice, NRS is more frequently used than VAS, even if less accurate, and it looks more clinically relevant to use a 4/10 NRS cutoff PASS point. When using a verbal scale, these cutoff points correspond to the border between mild and moderate, thus it may be considered that a treatment target should be no worse than mild pain. In our study, patients were included if they were experiencing pain at an intensity of more than 3 on an NRS from 0 to 10, corresponding to 30 mm on the VAS used in clinical trials. Our higher thresholds cannot therefore
be accounted for by a difference in the intensity of pain at inclusion. Our findings also show that the acceptability of OA pain is not homogeneous, with different levels of acceptability recorded for pain at rest and pain on movement. This aspect should be taken into account, as pain may differ considerably at rest and on movement, particularly when pain fluctuations are considered, and this may have an impact on analyses [10]. 4.2. Independent baseline determinants of PASS cutoff points Despite the similarity of the PASS thresholds determined here for KOA and HOA, these thresholds had different independent demographic and patient priority baseline determinants. We identified 5 independent baseline determinants for KOA PASS thresholds: age, sex, level of pain at rest, level of pain on movement, and the disability-related priority of the patient. By contrast, we identified only 2 independent baseline determinants for HOA PASS thresholds: level of pain at rest and BMI. Pain intensity levels at rest and on movement influenced the PASS thresholds for KOA, whereas only pain intensity at rest influenced the PASS threshold for HOA. Patient disability-related priority was an independent determinant factor only for the KOA PASS threshold. This suggests that even if pain levels do not differ significantly between HOA and KOA, the acceptability of HOA pain may be influenced by a larger number of independent factors than that of KOA pain. This finding differs from that for joint replacement acceptability, for which sex, age, and BMI do not seem to influence PASS cutoff points [7]. It also conflicts with the data obtained in randomized controlled studies of HOA and KOA, in which age, sex, and disease duration were found to have no effect on the frequency of PASS [21]. This may reflect the different coping strategies used by patients with HOA and KOA in real life [15]. It is also consistent with the findings recently reported by Escobar et al. [7], showing that patients undergoing hip replacement displayed a greater improvement in pain and functionality 1 year after transplantation, with a faster functional improvement, than patients undergoing knee replacement. In other studies, on ankylosing spondylitis for example [13], the duration of symptoms directly affects the PASS threshold. By contrast, in our study, OA duration did not influence the PASS threshold.
Ò
254
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
Table 5 Independent factors accounting for a lack of acceptability of osteoarthritis pain (according to patient acceptable symptom state cutoff point; knee osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Sex Male Female
1 (ref) 1.45
1.12–1.89
.004
Age, years <48–60 60–65 65–70 70–75 P75
1 (ref) 0.91 1.08 1.19 1.61
0.62–1.33 0.73–1.59 0.81–1.72 1.11–2.27
.617 .721 .369 .010
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 1.35 2.08 4.17
0.76–2.38 3.57–1.20 2.22–7.69
.314 .001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 1.52 3.85
0.69–3.33 1.75–8.33
.296 <.001
Most affected activities Household Outdoor Sports Leisure Social activities
1.09 1 (ref) 0.55 0.79 1.27
0.81–1.47
.578
0.35–0.86 0.53–1.19 0.73–2.22
.001 .268 .395
1 (ref) 0.68
0.46–1.02
.061
0.78
0.49–1.25
.298
0.97
0.66–1.43
.898
Disease-modifying treatment Level 1 analgesics Level 2 analgesics, with or without level 1 analgesics Level 1 or 2 analgesic with another drug Drugs plus nonpharmacological therapies
This may be because most patients had had symptoms for a long time and because it was difficult to define the exact time of onset of OA disease.
4.3. MCII cutoff points Our results for chronic OA pain show that MCII levels are similar for HOA and KOA, at rest and on movement, with an estimated 1-point decrease in the score attributed on an NRS running from 0 to 10. These findings are similar to those for other studies [6,16], but the improvement is smaller than that reported by Tubach et al. [22]: MCII of 19.9 mm for KOA and 15.3 mm for HOA on a VAS (0 to 100). This difference may reflect differences in the clinical setting: Pope et al. [16] demonstrated that MCII was lower in clinical practice than in randomized clinical trials. The likelihood of reaching the MCII was correlated with pain intensity, both at rest and on movement: the higher the pain intensity at baseline, the greater the proportion of patients reaching the MCII. These findings are consistent with those of other studies on OA [14,20] and arthritis [17] that have demonstrated that MCII depends on baseline pain score. In our study, age, sex, and the disability-related priority of the patients had no effect on MCII, whereas BMI was an independent determining factor for KOA, with obese patients less frequently achieving the MCII than patients with a normal BMI. These findings are consistent with those of other studies on MCII in functional index, showing that MCII is context-specific and not a fixed attribute [25]. 4.4. Limitations of the study This was an observational study, in general practice, with a short (7 days) follow-up period. The acceptability of OA pain may be different in chronic care and for longer durations, due to the higher frequency of adverse events or the larger proportion of patients lost to follow-up, with longer follow-up periods. For reasons of feasibility, only limited data were collected at the start of the study, principally demographic data, pain intensities, and the patients’ disability-related priorities. The analysis of PASS and MCII determinants was therefore limited to this restricted set of variables. We did not assess the patients’ psychological determinants, their expectations, or the extent to which they agreed with their physician on the management of the disease. It would be useful
Fig. 6. Proportions of patients reaching PASS at rest, on movement and in both situations, as a function of their disability-related priorities.
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256 Table 6 Independent factors accounting for improvement (according to the minimal clinically important improvement cutoff; hip osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 3.93 8.18 15.49
2.22–6.97 4.55–14.69 5.88–40.81
<.001 <.001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 3.73 3.21
1.53–9.10 1.28–8.08
.004 .013
Body mass index, kg/m2 <18.5 18.5–25 25–30 P30
0.22 1 (ref) 0.87 0.48
0.01–3.84
.299
0.57–1.34 0.29–0.81
.527 .006
Table 7 Independent factors accounting for improvement (according to minimal clinically important improvement cutoff) (knee osteoarthritis). Factor
Odds ratio
95% confidence interval
P value
Pain at rest, day 0 0–3 3–4 5–7 8–10
1 (ref) 3.17 5.08 4.43
2.11–4.75 3.40–7.59 2.52–7.81
<.001 <.001 <.001
Pain on movement, day 0 3–4 5–7 8–10
1 (ref) 2.50 2.38
1.51–4.13 1.38–4.09
.0004 .002
Body mass index, kg/m2 <18.5 18.5–25 25–30 P30
0.07 1 (ref) 1.06 0.80
0.01–0.76
.024
0.77–1.46 0.57–1.13
.718 .215
to collect such additional information in future studies to improve our understanding of the acceptability of chronic painful conditions such as OA because pain severity in OA has been shown to be associated with both demographic and psychological variables [4]. 4.5. Conclusions In primary care settings, patients with lower limb OA may be considered to be in an acceptable symptomatic state if they assign a score of <4 to their pain on an NRS from 0 to 10 (40 mm on a VAS) at rest and a score of <5 (50 mm on a VAS) on movement, corresponding to no worse than mild pain. The independent determinants of PASS in lower limb OA differ between HOA and KOA. For patients with KOA, the cutoff point for PASS is influenced by age, sex, and pain intensity at rest and on movement, whereas for patients with HOA, it is determined BMI and pain at rest. After pain management in daily practice, the cutoff for the MCII was a decrease of 1 point on an NRS from 0 to 10 ( 10 mm on a VAS). The odds of reaching MCII increased with pain intensities at rest and movement for both KOA and HOA and were also influenced by BMI in KOA. Thus, the PASS and MCII concepts are relevant to daily practice. They should be used in daily clinical practice, and not only in clinical trials, for a more personalized management of chronic pain states. PASS and MCII could be used in the future for the routine assessment of PROs for analgesic treatment.
255
Acknowledgments This study was developed with an unrestricted grant from SANOFI France. This article was prepared with the help of a medical writer, Dr. Christian Kempf, from CEGEDIM and Julie Sappa for English corrections. References [1] Altman RD, Asch D, Bloch G, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–49. [2] Arden NK, Kiran A, Judge A, Biant LC, Javaid MK, Murray DW, et al. What is a good patient reported outcome after total hip replacement? Osteoarthritis Cartilage 2011;19:155–62. [3] Beaton DE, Boers M, Wells GA. Many faces of the minimal clinically important difference (MCID): a literature review and directions for future research. Curr Opin Rheumatol 2002;14:109–14. [4] Creamer P, Lethbridge-Cejku M, Hochberg M. Determinants of pain severity in knee osteoarthritis: effect of demographic and psychosocial variables using 3 pain measures. J Rheumatol 1999;26:1785–92. [5] Dougados M, Moore A, Yu S, Gitton X. Evaluation of the patient acceptable symptom state in a pooled analysis of two multicentre, randomised, doubleblind, placebo-controlled studies evaluating lumiracoxib and celecoxib in patients with osteoarthritis. Arthritis Res Ther 2007;9:R11. [6] Ehrich EW, Davies GM, Watson DJ, Bolognese JA, Seidenberg BC, Bellamy N. Minimal perceptible clinical improvement with the Western Ontario and McMaster Universities osteoarthritis index questionnaire and global assessments in patients with osteoarthritis. J Rheumatol 2000;27:2635–41. [7] Escobar A, Gonzalez M, Quintana JM, Vrotsou K, Bilbao A, Herrera-Espiñeira C, Garcia-Perez L, Aizpuru F, Sarasqueta C. Patient acceptable symptom state and OMERACT-OARSI set of responder criteria in joint replacement. Identification of cut-off values. Osteoarthritis Cartilage 2011;20:87–92. [8] Heiberg T, Kvien TK, Mowinckel P, et al. Identification of disease activity and health status cut-off points for the symptom state acceptable to patients with rheumatoid arthritis. Ann Rheum Dis 2008;67:967–71. [9] Hochberg MC, Wohlreich M, Gaynor P, Hanna S, Risser R. Clinically relevant outcomes based on analysis of pooled data from 2 trials of duloxetine in patients with knee osteoarthritis. J Rheumatol 2012;39:352–8. [10] Hutchings A, Calloway M, Choy E, Hooper M, Hunter DJ, Jordan JM, Zhang Y, Baser O, Long S, Palmer L. The Longitudinal Examination of Arthritis Pain (LEAP) study: relationships between weekly fluctuations in patient-rated joint pain and other health outcomes. J Rheumatol 2007;34:2291–300. [11] Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989;10:407–15. [12] Kvien TK, Heiberg T, Hagen KB. Minimal clinically important improvement/ difference (MCII/MCID) and patient acceptable symptom state (PASS): what do these concepts mean? Ann Rheum Dis 2007;66:40–1. [13] Maksymowych WP, Gooch K, Dougados M, Wong RL, Chen N, Kupper H, van der Heijde D. Thresholds of patient-reported outcomes that define the patient acceptable symptom state in ankylosing spondylitis vary over time and by treatment and patient characteristics. Arthritis Care Res 2010;62:826–34. [14] Ornetti P, Dougados M, Paternotte S, Logeart I, Gossec L. Validation of a numerical rating scale to assess functional impairment in hip and knee osteoarthritis: comparison with the WOMAC function scale. Ann Rheum Dis 2011;70:740–6. [15] Perrot S, Poiraudeau S, Kabir M, Bertin P, Sichere P, Serrie A, Rannou F. Active or passive pain coping strategies in hip and knee osteoarthritis? Results of a national survey of 4719 patients in a primary care setting. Arthritis Rheum 2008;59:1555–62. [16] Pope JE, Khanna D, Norrie D, Ouimet JM. The minimally important difference for the health assessment questionnaire in rheumatoid arthritis clinical practice is smaller than in randomized controlled trials. J Rheumatol 2009;36:254–9. [17] Stauffer ME, Taylor SD, Watson DJ, Peloso PM, Morrison A. Definition of nonresponse to analgesic treatment of arthritic pain: an analytical literature review of the smallest detectable difference, the minimal detectable change, and the minimal clinically important difference on the pain visual analog scale. Int J Inflamm 2011;23:19–26. [18] Strand V, Boers M, Idzerda L, Kirwan JR, Kvien TK, Tugwell PS, Dougados M. It’s good to feel better but it’s better to feel good and even better to feel good as soon as possible for as long as possible. Response criteria and the importance of change at OMERACT 10. J Rheumatol 2011;38:1720–7. [19] Tashjian RZ, Deloach J, Porucznik CA, Powell AP. Minimal clinically important differences (MCID) and patient acceptable symptomatic state (PASS) for visual analog scales (VAS) measuring pain in patients treated for rotator cuff disease. J Shoulder Elbow Surg 2009;18:927–32. [20] Tubach F, Pham T, Skomsvoll JF, et al. Stability of the patient acceptable symptomatic state over time in outcome criteria in ankylosing spondylitis. Arthritis Rheum 2006;55:960–3. [21] Tubach F, Ravaud P, Baron G, Falissard B, Logeart I, Bellamy N, et al. Evaluation of clinically relevant states in patient-reported outcomes in knee and hip
256
Ò
S. Perrot, P. Bertin / PAIN 154 (2013) 248–256
osteoarthritis: the patient acceptable symptom state. Ann Rheum Dis 2005;64:34–7. [22] Tubach F, Ravaud P, Baron G, Falissard B, Logeart I, Bellamy N, et al. Evaluation of clinically relevant changes in patient-reported outcomes in knee and hip osteoarthritis: the minimal clinically important improvement. Ann Rheum Dis 2005;64:29–33. [23] Tubach F, Ravaud P, Beaton D, Boers M, Bombardier C, Felson DT, et al. Minimal clinically important improvement and patient acceptable symptom state for
subjective outcome measures in rheumatic disorders. J Rheumatol 2007;34:1188–93. [24] Tugwell P, Bombardier C, Buchanan WW, Goldsmith CH, Grace E, et al. The MACTAR Patient Preference Disability Questionnaire—an individualized functional priority approach for assessing improvement in physical disability in clinical trials in rheumatoid arthritis. J Rheumatol 1987;14:446–4451. [25] Wang YC, Hart DL, Stratford PW, Mioduski JE. Baseline dependency of minimal clinically important improvement. Phys Ther 2011;91:675–88.