Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Pyrrhic Victories Cause Untold damage
- Dr. Herbert H. Nehrlich (3 December 2004)
-
Pain and Osteoarthritis management
- Susan Margaret Oliver (12 December 2004)
-
Misleading conclusions in use of the minimal individual perceptible difference to interpret mean differences in change over placebo
- Florence Tubach, Bruno Giraudeau PhD, Philippe Ravaud MD PhD (28 December 2004)
-
Authors reply to Tubach et al.
- Jan M Bjordal, Anne Elisabeth Ljunggren, Atle Klovning, and Lars Slørdal (7 January 2005)
|
|
|||
|
Dr. Herbert H. Nehrlich, Private Practice Bribie Island, Australia 4507
Send response to journal:
|
Non-steroidal antiinflammatory drugs are potent weapons in the fight against pain when it has teamed up with its sibling inflammation. It must have occurred by now, even to the most dimwitted practitioner, that these drugs do nothing for the patient except win little repeated skirmishes against the emotion of pain. Each skirmish leaves more damage in its wake, although it looks like the battle has been won. There are a few doctors in various parts of the world who remember the experience of Dr. William Kaufman MD PhD who successfully used Vitamin B 3 (as Niacinamide) in the treatment of OA. The reaction of the patients to Kaufman's treatment is uniformly positive, the echo from the scientific community is condemnation and dire warnings about liver damage and other disasters . Dr. Abram Hoffer, who has also used both Niacinamide and Niacin in megadoses for Schizophrenia and related conditions, found no significant adverse effects in 50 years of use. Both physicians found the treatment effective and of no danger to the patient so one might be forgiven to ask why not everyone is using B 3. The answer is that it appears to be not scientific enough. Perhaps it will make a come-back as soon as access to food supplements for the lay person has been eliminated. Competing interests: None declared |
|||
|
|
|||
|
Susan Margaret Oliver, Chair of the Royal College of Nursing Rheumatology Forum and independent nurse specialist Litchdon Medical Centre. EX32 9BZ
Send response to journal:
|
Bjordal et als' systematic review studied the level of pain relief achieved using visual analogue scales following treatment with a non- steroidal anti-inflammatory or cyclo-oxygenas 2 inhibitors for those with osteoarthritis of the knee. They state in their introduction that guidelines highlight the need for both pharmacological and non- pharmacological interventions for optimal treatment of knee osteoarthritis yet this issue is not discussed as a limitation in reviewing simply the pharmacological options to treatment of knee osteoarthritis. Pain is a subjective and complex symptom and factors such as anxiety and stress also play a significant part in the individual's ability to cope with pain. These factors can be affected by the level of information provided about the disease as well as practical advice on simple pain management principles or practical non-pharmacological advice on the use of appropriate footwear etc. It is therefore hardly surprising that individuals treated simply with an NSAID or COX2 only achieved some short term pain relief that was not sustained over time. Competing interests: I am member of Standards for Arthritis Relief (SOAR) and ARthritis Action Group supported by an educational grant from Pfizer. |
|||
|
|
|||
|
Florence Tubach, MD INSERM E0357, Dt d’Epidémiologie, Biostatistique et Recherche Clinique, Hôpital Bichat 75018 Paris, Bruno Giraudeau PhD, Philippe Ravaud MD PhD
Send response to journal:
|
The meta-analysis of randomised controlled trials assessing the efficacy of nonsteroidal anti-inflammatory drugs in osteaoarthritic knee pain performed by Jan Magnus Bjordal et al.1 is technically well done, but the authors’ conclusions are misleading. They assert that the mean change in pain, as measured on a visual analogue scale (VAS), over placebo was 10.1 mm [7.4 to 12.8] and claim a non relevant difference, since the minimal clinically perceptible difference was 9.7 mm2. Actually, the authors mixed up the group and the individual level interpretations. Indeed, 9.7 is the perceptible difference at the individual level (i.e., a change lower than 9.7 mm would not be perceived by the patient). However, the 10.1 mm estimate assessed by the authors makes sense only at the group level. To better understand this multi-level interpretation, we performed some calculations classifying patients as improved or not if they achieved a change greater than 10.1 mm. Thus, hypothesizing that in the placebo group, the change equals 5+20 mm, 40.1% of the placebo group patients would be improved (under the reasonable assumption of a normal distribution of change in pain on a VAS). A treatment effect of 10.1 mm leads to a mean change of 15.1 mm in the experimental group, and thus 60.1% of patients would be improved. The number of patients needed to treat is then estimated at 5.0. In the same way, if the difference in the placebo group was 20±20 mm, 69.2% of patients would be improved in this group and 84.3% in the experimental group, and the number needed to treat would then be 6.6. These examples illustrate that using the minimal clinically perceptible difference to interpret changes at the group level is inadequate: a small variation at the group level does not mean no clinically relevant change in the individuals of the group. 1. Bjordal JM, Ljunggren AE, Klovning A, Slordal L. Non-steroidal anti-inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials. BMJ 2004;329:1317-XX? 2. 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(11):2635-41. Competing interests: None declared |
|||
|
|
|||
|
Jan M Bjordal, Research Fellow ISF, University of Bergen, 5018 Bergen, Norway, Anne Elisabeth Ljunggren, Atle Klovning, and Lars Slørdal
Send response to journal:
|
Drs Tubach, Giraudeau and Ravaud claim that we misinterpret our data(1) by confusing group and individual responses and thereby reach a misleading conclusion. To underline their opinion, they make hypothetical number needed to treat (NNT) calculations of 5 and 6.6 in a best-case scenario for NSAIDs, based on the highest (biased) weighted mean difference (WMD) and the lowest possible benchmark for individual response. Patient response on a group level can be expressed as effect size/value (ES) or WMD on a 100 mm visual analogue scale (VAS), while NNT values tell us on an individual level how likely patients are to respond. In our meta-analysis, we decided to use ES and WMD for two reasons; because of the underlying data presentation in single studies, and because universal standards exist for the interpretation of effect for these parameters(2). In a subgroup analysis, we found that patient selection bias systematically inflated results in a subgroup of trials. Additional (unpublished) analyses confirm this finding. For group (or distribution- based) responses, the unbiased ES of NSAIDs in our meta-analysis was clearly poor at 0.23 [95% CI: 0.16-0.31]2, and unbiased WMD was 5.9mm [95% CI: 3.8-7.9] on VAS. Whether to use group (distribution) or individual (anchor-based) approaches in interpretation of results, may be open to discussion. However, they have a near-linear relationship (3). An additional problem for individual approaches is to agree upon a clinically relevant benchmark for dichotomising treatment success or failure. For knee osteoarthritis, empirically developed subjective outcomes range from mean values of 9.7 – 19.9mm, for minimally perceptible clinical improvement (MPCI)(4), minimally clinically important improvement (MCII)(5), and minimal clinically important difference (MCID)(6), respectively. If these thresholds are used as benchmarks for treatment success together with the unbiased WMD, NNT-values will range between 9 and 20. Tubach et al. also attempt to make a point of the fact that single patients actually may experience benefit. In our opinion, it is more relevant to question if patients should be put at risk for adverse effects from oral NSAIDs, when less than 20% of them would notice if oral NSAIDs were replaced by a placebo. We stand firmly by our conclusions from the meta-analysis(1). In our opinion, the available data clearly demonstrate a lack of clinically relevant effects from NSAIDs in knee osteoarthritis pain. Jan M. Bjordal Anne Elisabeth Ljunggren Atle Klovning Lars Slørdal References: 1. Bjordal JM, Ljunggren AE, Klovning A, Slordal L. Non-steroidal anti-inflammatory drugs, including cyclo-oxygenase-2 inhibitors, in osteoarthritic knee pain: meta-analysis of randomised placebo controlled trials. BMJ 2004;329(7478):1317-. 2. Cohen J. Statistical power analysis for behavioural sciences. Academic Press, New York, USA 1977. 3. Norman GR, Sridhar FG, Guyatt GH, Walter SD. Relation of distribution- and anchor-based approaches in interpretation of changes in health-related quality of life. Med Care 2001;39(10):1039-47. 4. 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(11):2635-41. 5. Tubach F, Ravaud P, Baron G, 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(1):29-33. 6. Angst F, Aeschlimann A, Michel BA, Stucki G. Minimal clinically important rehabilitation effects in patients with osteoarthritis of the lower extremities. J Rheumatol 2002;29(1):131-8. Competing interests: None declared |
|||