Doi:10.1016/s0140-6736(00)04161-

Research letters
George A C Murrell, Judie R Walton Rotator cuff tears account for almost 50% of major shoulder second set consisted of the first 100 patients who had injuries but are sometimes difficult to diagnose. To aid shoulder pathology not involving a rotator cuff tear (no tear diagnosis, we did a prospective study, comparing results of 23 clinical tests from 400 patients with and without rotator cuff Most clinical shoulder tests could not distinguish between tears. Three simple tests were predictive for rotator cuff tear: the two groups. However, the tests for three clinical features supraspinatus weakness, weakness in external rotation, and were more positive in patients with rotator cuff tears than in impingement. When all three were positive, or if two tests were the no tear set (p<0·0001) and were predictive for this positive and the patient was aged 60 or older, the individual had disorder. These features were: supraspinatus weakness; a 98% chance of having a rotator cuff tear; combined absence of weakness in external rotation; and impingement (in internal these features excluded this diagnosis.
rotation, in external rotation, or in both directions). Because Rotator cuff tears are a common source of shoulder pain, of its high specificity (98%), the drop-arm sign was also especially in people aged 60 years and older.1 To improve the predictive for rotator cuff tear, but its sensitivity was only 10% clinical diagnosis of rotator cuff tears, we did two studies. In the first study we aimed to find out whether any of 23 clinical At least 15% of people with rotator cuff tear have additional tests commonly used in shoulder assessment were predictive shoulder pathology. Hence, we designed the second study to for rotator cuff tear. The purpose of the second study was to verify our findings on a typical patient population. We verify the predictability of the first. All patients were referred compared the test results of all 200 patients who presented to the practice of an orthopaedic surgeon (GM) at St George with a rotator cuff tear (whether or not they had additional Private Hospital, Sydney, Australia, between May, 1996, and shoulder pathology) with 200 in the no tear set. December, 1999, and had shoulder injuries sufficient to We established the prevalence of rotator cuff tear for all patients for each 10-year age-group. RCT prevalence (%= In the first study, we systematically assessed a large group RCT/[RCT+NT]) with methods described by Sackett and of patients with 23 commonly used shoulder tests. Individuals colleagues,4 and calculated the post-test probabilities for were examined for wasting, tenderness, and signs (the drop- rotator cuff tear based on the clinical test results and age- arm sign, O’Brien’s sign, and impingement),2 as well as active and passive ranges of motion, and shoulder strength.3 We The results of the second study confirmed the results of the assessed strength by manual muscle tests on a scale of 0–5, first with the same degree of significance (p<0·0001).
with an additional score of 4·5 to show near to normal Furthermore, rotator cuff tear frequency increased linearly strength. Any score less than 5 was interpreted as indicating with age from the third decade onwards—for example, rising weakness, providing that there was no concomitant loss of from 33% in the 40s to 55% in the 50s (figure). strength during internal rotation. The patients were grouped, Patients who present with shoulder pain, and who test on the basis of their arthroscopic findings, into two sets. One positive for supraspinatus weakness, weakness in external set consisted of the first 100 individuals diagnosed with a rotation, and impingement, have a 98% chance of rotator cuff partial or full-thickness rotator cuff tear of any size and no tear (table).5 If they test positive for any two of these clinical other major shoulder pathology (rotator cuff tear set). The features and are aged 60 years or older, they still have 98%chance of a rotator cuff tear. If only one of the three tests is positive, the clinical result is indeterminate and imaging is needed. Any patient with a positive drop-arm sign also has a 98% chance of rotator cuff tear. If none of these clinicalfeatures are present, the chance of having rotator cuff tear fallsto only 5% and, for practical purposes, can be ruled out in the absence of supervening pathology or other degeneration.
0–19 20–29 30–39 40–49 50–59 60–69 70–79 *Supraspinatus weakness, weakness in external rotation, and positive impingement sign or signs. †Age-adjusted. ‡Approximation based on a fixed total count and an Fdistribution.5 Rate of rotator cuff tear per 10-year age-group Post-test probabilities for rotator cuff tear in different THE LANCET • Vol 357 • March 10, 2001 For personal use only. Reproduce with permission from The Lancet Publishing Group.
The predictive power of the combined clinical tests is similar to the best values for magnetic resonance imaging and ultrasonography. Moreover, the clinical tests are easy and We thank the patients and medical practitioners participating in this study,and David L Sackett and George Brown for statistical advice. St George Hospital/South Eastern Sydney Area Health Service, St GeorgePrivate Hospital/Health Care of Australia, and Mitek Australia Ltdsupported the research.
1Norwood LA, Barrack R, Jacobson KE. Clinical presentation of complete tears of the rotator cuff. J Bone Joint Surg Am 1989; 71:
499–505.
Reider B. The orthopaedic physical examination. Philadelphia: WB Saunders, 1999: 49–53. Hawkins RJ, Bokor DJ. Clinical evaluation of shoulder problems. In:Rockwood JCA, Matsen IFA, eds. The shoulder, vol 1, 1st edn.
Philadelphia: WB Saunders, 1990: 149–77.
Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence- based medicine: how to practice and teach EBM. New York: ChurchillLivingstone, 1997: 228–34.
Leemis LM, Trivedi KS. A comparison of approximate interval estimators for the Bernouli parameter. Am Stat 1996; 50: 63–68.
Figure 1: Parasite prevalence before and after albendazole Orthopaedic Research Institute, University of New South Wales, St George Hospital Campus, Sydney, 2217, Australia At a single dose of 400 mg, albendazole is known to be Correspondence to: Prof George A C Murrell highly effective against hookworm infection.1,4 However, since previous studies have shown the need for multiple doses ofalbendazole to treat the cohabiting enteric pathogen Giardia,the initial community therapy was designed to reduce prevalence of Giardia as well as hookworm; thus a 5-day treatment regimen of 400 mg/day was given in 1993.4 Since1996, the drug has been given as a single dose of 400 mg to all children older than 1 year of age every 6 months and to allconsenting adults annually (excluding pregnant women, R C A Thompson, J A Reynoldson, S C Garrow, J S McCarthy, pregnancy status). In 1993, the pretreatment enteric parasitestatus of the whole community was calculated, and, before Hookworm (Ancylostoma duodenale) and other enteric parasites
subsequent mass treatments, volunteer subsets of the such as Giardia and Hymenolepis are common in Aboriginal
population in the age range 2–23 years (with three adults communities in northern Australia, and their presence is aged 48, 56, and 72 years in the most recent survey) were associated with iron deficiency, anaemia, and failure to thrive.
assessed. Over all the trials, the proportions of individuals We report the outcome of a sustained, community-based controlprogramme that used regular albendazole in one isolated community. Whereas hookworm has been effectively controlled
by the programme, no sustained effect on the prevalence of
Giardia and Hymenolepis was seen; the control of these
parasites will depend on improvements in health education. Thisprogramme might serve as a model for community-based orpopulation-based drug treatment programmes elsewhere.
Hookworm infection is a major cause of morbidity in endemic areas throughout the world, manifesting clinically asiron-deficiency anaemia.1 This is the case in Aboriginalcommunities in northern Australia where iron deficiency andanaemia are common in children and women.2 Although there is a paucity of accurate prevalence data available for thisregion, a detailed survey undertaken in 1992 in a remote, coastal community of about 350 people found that the overall prevalence of hookworm (Ancylostoma duodenale) infectionwas 77%, reaching 93% in children 5–14 years of age.2Previous attempts to control hookworm in this community,which had combined pyrantel treatment with environmental and health-education strategies, were hampered by resistance to pyrantel.3 In this report, we document the outcome of asustained, community-based control programme that usedregular albendazole, combined with continued health education and environmental management, from 1993 to 1999. In addition, there has been regular surveillance of hookworm status in the population, the lack of whichcontributed to the failure of previous campaigns to eradicate THE LANCET • Vol 357 • March 10, 2001 For personal use only. Reproduce with permission from The Lancet Publishing Group.

Source: http://www.sportsclinicnq.com.au/assets/files/murrellcufftear.pdf