Dear Editor,

We thank Dr Chirinos for his interest in our article and appreciate his comments [1].

We would disagree, however, with his suggestion that paired t-test is only appropriately used to analyze repeated measurements before and after an intervention in a single sample population. In our study the statistical analysis by a paired t-test was decided before data collection and based on the study design of cases and controls being closely matched for five variables (i.e. sex, age, height, mean peripheral blood pressure and heart rate).

We are very happy to provide the probability values for the difference of the augmentation index between rheumatoid cases and healthy controls, using paired/ unpaired statistical tests for normally and not normally distributed data. These were 0.0284, 0.0245, 0.0422 and 0.0432 for paired t-test, Wilcoxon (matched pairs) test, unpaired t-test and Mann -Whitney test, respectively.

The difference remains statistically significant and we would therefore maintain that our study shows that arterial stiffness is indeed increased in patients with rheumatoid arthritis, a finding which has recently been confirmed both in patients with rheumatoid arthritis [2] and those with systemic vasculitis [3].

References

(1). Chirinos AJ. Is arterial stiffness increased in rheumatoid arthritis? Ann Rheum Dis 2004; (eLetter; 24 June).

(2). Wong M, Toh L, Wilson A, Rowley K, Karschimkus C, Prior D, Romas E, Clemens L, Dragicevic C, Harianto H, Wicks I, McColl G, Best J, Jenkins A. Reduced arterial elasticity in rheumatoid arthritis and the relationship to vascular disease risk factors and inflammation. Arthritis Rheum 2003;48:81-89.

(3). Booth AD, Wallace S, McEniery CM, Yasmin, Brown J, Jayne DR, Wilkinson IB. Inflammation and arterial stiffness in systemic vasculitis: a model of vascular inflammation. Arthritis Rheum 2004;50:581-588.

Dear Editor,

We read with great interest the article by Allanore et al, and commend on their efforts in studying this important issue[1]. We would be most grateful if the authors can help to clarify a few points.

Mean Deviation (MD) is the average of the numbers on the total deviation plot with each value weighted according to the magnitude of the normal range at that point. It signifies the overall severity of the field loss. Thus, a MD of -2 dB depression may indicate a 2 dB depression everywhere in the field (not necessarily a glaucomatous field loss) or a depression of 4 dB over half of the field (more specific for a glaucomatous process). According to one of the commonly employed criteria from Anderson [2] glaucomatous visual field defects may be defined by (1) three or more adjacent points in an expected location of the central 24o field, on the same side of the horizontal meridian, that have p <_5 on="on" the="the" pattern="pattern" deviation="deviation" pd="pd" plot="plot" one="one" of="of" which="which" must="must" have="have" p1.="p1." _2="_2" glaucoma="glaucoma" hemifield="hemifield" test="test" ght="ght" being="being" outside="outside" normal="normal" limits.="limits." _3="_3" corrected="corrected" standard="standard" cpsd="cpsd" with="with" p="p" _5.="_5." as="as" and="and" are="are" routine="routine" printout="printout" statistics="statistics" humphrey="humphrey" sita="sita" _24="_24" visual="visual" it="it" would="would" be="be" interest="interest" if="if" authors="authors" can="can" provide="provide" these="these" indices="indices" for="for" their="their" subjects="subjects" will="will" helpful="helpful" in="in" confirming="confirming" that="that" field="field" defect="defect" was="was" indeed="indeed" glaucomatous.="glaucomatous."/> In the absence of progression of glaucomatous visual field loss, a patient having such field loss at a normal intraocular pressure (IOP) is at the most a normal tension glaucoma suspect. The authors may like to report any progression of glaucomatous visual field loss in the future, which will truly define normal tension glaucoma.

We concur with the author that cup-disc ratio (CDR), if measured vertically and > 0.7 in the absence of an extremely large or tilted cup, usually suggest glaucomatous optic neuropathy. A CDR of > 0.3, however, will have much less bearing in indicating a glaucomatous process. Studies have shown that vertical CDR must be interpreted with the vertical disc diameter (VDD) [3]. For example, in normal population (glaucoma eyes excluded) with VDD of 1.9 mm, the 50th percentile of vertical CDR is 0.55, with a 95th percentile at 0.74.

In addition, nowadays, the neuroretinal rim is regarded as more important than the cup or CDR. The cardinal feature of glaucomatous optic neuropathy is a loss of tissue from inner edge of the rim [4]. Other features include rim notching, hemorrhage crossing the rim, undercutting of rim, and asymmetry of rim between eyes, etc [4]. The authors may like to provide these data in support of a glaucomatous optic nerve damage.

Finally, both eyes of each subject were used in analysis. Any effect may be double-represented in the study population. This may be overcome by including only one eye of each patient, or if both eyes are to be included, using statistical means such as a generalized estimating equation [5].

Reference

(1). Allanore Y, Parc C, Monnet D, Brezin AP, Kahan A. Increased prevalence of ocular glaucomatous abnormalities in systemic sclerosis. Ann Rheum Dis. 2004 Oct; 63(10):1276-8.

(2). Anderson DR, Patella VM. Automated Static Perimetry. Mosby. 2nd edition. 1998

(3). Crowston JG, Hopley CR, Healey PR, Lee A, Mitchell P; Blue Mountains Eye Study. The effect of optic disc diameter on vertical cup to disc ratio percentiles in a population based cohort: the Blue Mountains Eye Study. Br J Ophthalmol. 2004 Jun;88(6):766-70.

(4). South East Asia Glaucoma Interest Group (SEAGIG). Asia Pacific Glaucoma Guidelines. 2004. p 21

(5). Katz J, Zeger S, Liang KY. Appropriate statistical methods to account for similarities in binary outcomes between fellow eyes. Invest Ophthalmol Vis Sci. 1994 Apr;35(5):2461-5.

Dear Editor,

I was surprised to find in the current issue of Annals of the Rheumatic Diseases that the lead editorial (“Leader” in British English) was written by the co-authors of one of the manuscripts reviewed in the same editorial (1,2).

Traditionally, readers of the peer-reviewed biomedical literature have come to expect that editorialists bring an unbiased perspective to the papers or topics being reviewed. It would seem to be a fundamental violation of that implicit trust to permit authors to report their own findings and then concurrently to submit an editorial which compares their work with the other papers in the field.

An example of the bias that may result from this apparent conflict of interest is that the importance of the findings may be inflated. Here, for example, the editorial is entitled “A historic issue of the Annals: three papers examine paracetamol in osteoarthritis,” thus implying that these are epochal papers of profound historic gravity; close examination of the material, however, leaves one substantially less enthusiastic that major breakthroughs have occurred. Moreover, although in the body of the editorial the authors mention the extremely short trial periods employed by each of the positive paracetamol studies, they conclude that “…the aggregated research data still support paracetamol as being more effective than placebo in relieving pain of large joint OA.” As OA is a chronic disease characterized by recurrent pain over years, findings of efficacy at one week and 3 months can hardly be considered compelling evidence of efficacy in relieving the pain of OA.

The point is not that paracetamol is ineffective; rather, its efficacy for chronic pain relief remains unstudied and unclear.

Reference

(1). Neame R, Zhang W, Doherty M: A historic issue of the Annals: three papers examine paracetamol in osteoarthritis. Ann Rheum Dis 2004; 63:897-900. 2. Zhang W, Jones A, Doherty M: Does paracetamol (acetaminophen) reduce the pain of osteoarthritis?: a meta-analysis of randomized controlled trials. Ann Rheum Dis 2004; 63:901-907.

Dear Editor,

We read with interest the article by Shanahan et al [1] on suprascapular nerve blocks and the accompanying editorial by Hall and Buchbinder [2]. It is gratifying to know that in this context the indirect method of needle placement produces a similar outcome to the radiologically-guided method. However, we would like to point out an important methodological flaw in the study.

The patients who received injections via the landmark approach were injected with 40mg of methylprednisolone and 10 mls of 0.5% bupivacaine, whereas those who underwent CT-guided injection received 40mg of methylprednisolone and 3 mls of 0.5% bupivacaine. Previous studies have shown that suprascapular nerve blocks produce effective analgesia for a variety of painful shoulder conditions, including frozen shoulder [3], rotator cuff lesions [4], and rheumatoid arthritis [5]. Fundamentally, however, in a double-blind study of 58 rheumatoid shoulders, suprascapular nerve block with bupivacaine alone has been shown to be as effective as bupivacaine plus methylprednisolone at improving pain, stiffness, and range of movement [6]. The authors of this study concluded that the addition of methylprednisolone conferred no additional benefit.

Thus, if one concludes from this that the local anaesthetic is the active ingredient, Shanahan’s two patient groups were given different doses of the same drug (10 mls compared with 3 mls). Consequently the study may have underestimated the effect of the CT-guided approach.

References

(1). Shanahan EM, Smith MD, Wetherall M et al. Suprascapular nerve block in chronic shoulder pain: are the radiologists better? Ann Rheum Dis 2004;63:1035-1040

(2). Hall S, Buchbinder R. Do imaging methods that guide needle placement improve outcome? Ann Rheum Dis 2004;63:1007-1008

(3). Dahan TH, Fortin L, et al. Double blind randomized clinical trial examining the efficacy of bupivacaine suprascapular nerve blocks in frozen shoulder. J Rheumatol. 2000 Jun;27(6):1464-9.

(4). Vecchio PC, Adebajo AO, Hazleman BL. Suprascapular nerve block for persistent rotator cuff lesions. J Rheumatol. 1993 Mar;20(3):453-5.

(5). Emery P, Bowman S, et al. Suprascapular nerve block for chronic shoulder pain in rheumatoid arthritis. BMJ. 1989 Oct 28;299(6707):1079-80.

(6). Gado K, Emery P. Modified suprascapular nerve block with bupivacaine alone effectively controls chronic shoulder pain in patients with rheumatoid arthritis. Ann Rheum Dis. 1993 Mar;52(3):215-8.