TOTAL population studies are the only certain way of gaining reliable information on the prevalence of insidious conditions like chronic simple glaucoma. The sine qua non of a total population study is exact knowledge of the composition of the population. In the Rhondda valley the Medical Research Council has laboriously carried out a census so that the exact number of persons and their age and sex are known for this area. Over the past 15 years the M.R.C. has earned the co-operation of this population by its efforts in detecting miners' pneumoconiosis, pulmonary tuberculosis, and other diseases. Such a situation was admirably suited for a total population study of the prevalence of glaucoma, and in the summer of 1963 the authors carried out such a study. To our knowledge the only previous similar study is that done by Str6mberg (1962) at Skovde in Sweden.The aims of our study were fourfold:(1) Collection of reliable data on the distribution of intra-ocular pressure in an unselected population when measured with both indentation and applanation tonometers.(2) Selection from this population of a group of persons who by conventional ophthalmological opinion were glaucoma suspects, so as to study these people and attempt a randomized trial of medical therapy.(3) The detection of unknown cases of glaucoma and estimation of the true prevalence of glaucoma of all types.(4) Tabulation of the extent of eye disease in an unselected population. Methods of Survey (1) The PopulationThree villages were chosen as containing a population that was considered to be of manageable size. All were covered by the M.R.C. census. The age and sex composition of the population is shown in Table I (opposite).In view of the difficulty of examining the very old, it was decided that persons above 75 years of age would not be examined. The lower limit was set at 40 years. There were 4,608 persons within these age limits, of whom we examined 4,231-91-9 per cent. The 8 1 per cent. who were not examined were interviewed and if they gave a history of treatment for eye disease details were sought at the medical centre involved. We feel that the coverage of this survey is adequate and better than most studies of this size. A team of professional *

In a short paper such as this, it would be a mistake to attempt a comprehensive review of our knowledge of the occurrence of strabismus and those characteristics of populations which may lead to variations in its incidence. To do so would subject us all to mental dyspepsia from a surfeit of information which none of us would have time to consider critically. Instead, I propose to take certain aspects of the epidemiology of strabismus, and to illustrate these by some observations made in Cardiff, with occasional data from other sources.The sort of questions we can hope to answer by epidemiological studies are those concerned with the prevalence of squint:How common is it?Does the prevalence vary from one community to another, and if so, can we identify any characteristics of these communities which may explain the difference?How do new cases arise, and at what ages?The first of these questions can be answered reasonably accurately, but the others present difficulties. The common snag in population research is that a study of previously published work reveals that sampling methods and criteria used in definition are either not stated or differ so much from one another that it would be most unwise to draw any conclusions from comparisons of prevalence. As a result, it becomes necessary to use mainly internal comparisons, and in particular the techniques of comparing a group of strabismic patients with a group of non-strabismic patients drawn at random from the original population, and examined by the same techniques. A number of the comparisons I shall present comes into this category. Material and techniquesA brief description of the nature of our population and our examination procedures and definitions is clearly the first step in describing some of our results.Our sample consisted of all the children born in the City of Cardiff between January I and December 3I of one year. The details of the 4,832 children born within these limits were obtained from the birth records maintained by the Medical Officer of Health, and the children were examined wherever possible at school during their first year by a research orthoptist who carried out a cover test at both distant and near fixation and estimated visual acuity. Those children not available during visits to school were sought out and examined where and when possible, and the coverage thus included private schools and schools for handicapped children.This technique, with a population largely captive in schools, should be and proved to be capable of giving a very high coverage of the population under study. The final total of children screened was 4,784, almost exactly 99 per cent. of the total sample.From the results of this initial examination, a selection was made on the basis of cover test results. Those with certain cover test responses were asked to return for a much more detailed orthoptic investigation and the taking of a full history including a number of questions related to the social copyright.

No abstract

The techniques of epidemiology originated in the study of the factors determining the spread ofinfectious disease. To-day, the term can be taken to mean, in a broader sense, the elucidation by the study of (3) Most surveys have relied on tonometry as the first step in identifying subjects.Schwartz (i965) pointed out that small systematic errors, such as can easily result from minor differences in tonometric technique, can produce quite large variations in "positive" screening tests.When those surveys which are suspect for epidemiological purposes are eliminated, few remain. Those which may be considered acceptable place the prevalence of clinical glaucoma, defined by the presence of tension over 21 mm.Hg, cupping of the disc, and field loss, at figures for Great Britain rather below those already quoted-for example, Hollows and Graham (i 966a) 0-47 per cent., Bankes, Perkins, Tsolakis, and Wright (I968) 0 7 per cent.These surveys, all designed to produce as accurate estimates of prevalence as possible by the exercise of care in sampling, give the most reliable assessments at present available

and exercise). The primary outcome measure was hip-related quality of life using the patient-reported International Hip Outcome Tool (iHOT-33) at 12 months (score 0-100, higher scores indicate better quality of life, and the minimal clinical important difference is 6.1). Secondary outcomes included generic health-related quality of life (EQ5D5L, SF12), adverse events, and cost-effectiveness. Primary analysis compared the differences in iHOT-33 scores at 12 months between groups, by intention to treat. Cost-effectiveness analysis took an NHS and personal social services perspective. Results: Of 6028 patients attending hip clinics at participating sites, 648 patients were eligible and 348 were randomised (171 to HA; 177 to PHT). There were no important differences between the groups in baseline characteristics. The average time to surgery was 132 days (SD 71) versus 47 days (SD 52) to PHT. 92.5% were followed-up for the primary outcome at 12 months. Baseline mean and standard deviation in iHOT-33 scores were 39.2 (SD 21) and 35.6 (SD 18) in the surgery and PHT groups, and 58.8 (SD 27) and 49.7 (SD 25) at 12 months, respectively. On average, patients in both groups improved over 12 months, and the mean iHOT-33 score increased more in those allocated to HA than to PHT, with a mean difference of 6.8 points (95% CI 1.7,12.0 p ¼ 0.009) in favour of surgery. Muscle soreness was common in both groups; one HA patient developed an infection requiring further surgery. Mean overall costs were £3713 for HA and £1283 for PHT. Conclusions: Treatment of patients with FAI syndrome with a strategy of either physiotherapy-led best conservative care or arthroscopic hip surgery led to improved hip-related quality of life. At 12 month followup that improvement was significantly greater in those allocated to surgery than in those allocated to conservative care. Hip arthroscopy was safe, but is associated with higher overall cost than conservative care over 12 months.