SummaryEighty-nine and a half per cent of the population of a general practice over the age of 49 years were screened for glaucoma and high risk ocular hypertension requiring treatment. Screening took place using semi-automated intraocular pressure and visual field equipment operated by non-ophthalmologicaUy trained staff. An experi enced ophthalmologist examined aU patients in a single blind manner to reduce false negatives to a minimum. Patients suspected of requiring treatment on the grounds of raised intraocular pressure, abnormal visual fields or suspicious optic discs were subsequently examined in a hospital clinic. Treatment criteria, as commonly prac ticed, were carefully defined and the sensitivities and specificities of the methods of screening used were calculated. One and three tenths per cent of the practice popula tion were known to be receiving treatment prior to the study and a further 1.4%were found to require treatment after screening.The sensitivity and specificity of the non-contact tonometer were 91.7% and 95.6% respectively with a predictive power of 22.5% for a positive result. The mean time taken to perform the test in both eyes was two minutes. Seventy per cent of the patients with pressures over 22 mmHg in both eyes on screening were found to require treatment. The routine use of the field screener did not increase either the sensitivity or specificity of the screening process but its use in cases with raised intraocular pressure is advised to indicate the degree of urgency of the referral.An algorithm based on the results of the study is suggested when planning the use of semi-automated equipment to screen for ocular disease related to raised intra ocular pressure.Chronic Simple Glaucoma (CSG) is a com mon disease of middle age and the elderly, prevalence rising from 0.5-1 % of the popula tion over the age of 401,2 to 6.6% over 75.3 It is responsible for up to 16.8% of registrable blindness in the population of the United Kingdom aged over 65 years.4CSG is characterised by a triad of physical signs: raised intraocular pressure (lOP),

Nottingham SummaryA survey of optometrists practising in Nottingham was undertaken to identify their methods of screening for glaucoma in the community. There was found to be great variation in all parameters examined including referral criteria. Although all of the respondents screened for glaucoma, 50% appeared not to be aware of subtle optic disc signs of the disease, 8% never measured intra-ocular pressures and 19% never performed visual field analysis. In a general practice population age 50 and over, 73% of patients had visited their optician within the last two years, 67% knew glau coma was an eye disease, but only 15% remembered being screened for glaucoma.Chronic simple glaucoma (CSG) is a major cause of blindness in the western world,l is asymptomatic in most cases until extensive loss of peripheral visual field occurs2 and the prevalence increases with age from 0.5-1 % of the population over 40, 3 rising to over 6% in the 75+ age group.4

Nottingham SummaryThe cost of detecting a case of glaucoma in a community based screening programme was calculated using data from a pilot study on a population aged 50 and over.The cost per screen was estimated at £3.35 with a total cost of £311 per case detected. These estimates include the cost of both the algorithmically based screen ing programme and the hospital assessment of true and false positives.These costs are considered to compare favourably with the potential costs incurred by paying optometrists to screen high risk groups whilst providing the opportunity to detect over twice the number of cases of occult glaucoma.

SummaryData from a glaucoma screening study involving 88.5% of the population age 50 and over of a single handed general practitioner were reanalysed to determine the effect of altering the protocol for intraocular pressure assessment and the effect of changing the referral threshold.The predictive power of the Keeler Pulsair noncontact tonometer was found to decrease from 22.5% at four pulses per eye to 12.3% when only one pulse per eye was used, with a reduction of sensitivity from 91. 7% to 75%.The sensitivity of the same device fell from 91.7% if all patients with an lOP >21 mmHg were deemed as having a positive screen, to 41.6% when only patients with an lOP >26 mmHg were considered for referral.To create a balance between high sensitivity and acceptable predictive power of a positive result in a population where 50% of glaucoma sufferers are known prior to screening, we advise that four pulses per eye should be used with an lOP of > 22 mmHg used as the significant finding indicating that the patient required referral.Tonometry alone has a poor reputation as a screening tool in the detection of glaucoma, as witnessed by a number' of epidemiological studies.l,2,) These studies used a definition of glaucoma which included a demonstration of field loss, although only one of the studies per formed visual field analysis in all of the popu lation studied and this study was limited to those individuals aged between 55 and 69 years.) Visual field loss is now known to be a rela tively late feature of glaucomatous damage to the optic nerve4 and treatment is often advis able in the absence of demonstrable field loss if other risk factors are present.5,6,7In Great Britain, many optometrists screen for glaucoma but a recent study has found that, in one major city, 50% of optometrists will not refer a patient for ophthalmological assessment unless the lOP is >24 mmHg.H Seventy-one per cent of the optometrists used non-contact tonometers (NeT) but 35% of these were measuring the lOP using only one or two pulses per eye.We have previously reported the results of a pilot study on the use of semi-automated equipment to screen for glaucoma in the com munity utilising non-ophthalmically trained staff.9,lO This study, examining 88.5% of a population aged 50 and over, used the fol-

Suprathreshold field screeners are in common use for the detection of glaucomatous field loss. The predictive power of a positive result (PP+) depends on the sensitivity and specificity ofthe screener in the population in which it is to be used. Using data from 755 normal individuals (1510 eyes), we calculated the PP+ of the Henson CFS2000 screening programme for a population aged 50 and over. 4*3% of normal eyes failed the screening programme. Ignoring one or two misses on the screening programme immediately adjacent to the disc reduced this figure to 1*3% and significantly improved the PP+ of the programme. Calculations ofthe PP+ at increasing glaucoma prevalence levels indicates this to be particularly relevant at low levels such as those encountered when screening middle aged and elderly populations. Optometrists should perform routine field analysis when screening for glaucoma provided they adhere to strict protocols.indicates a true positive identification of the disease process and is related to both sensitivity and specificity.To calculate the predictive power of the screening programme on the Henson CFS2000 in a population, the following data are required from that population: (a) the incidence of screen failures in normal persons; (b) the prevalence of field defects from other causes; (c) the prevalence of glaucomatous field loss; (d) the false negative rate on screening eyes with field loss from glaucoma.If (c) Therefore in order to estimate the predictive power of the Henson CFS2000 screening programme in the field we analysed the data on normal persons and patients with non-glaucomatous field defects identified on screening a general practice population aged 50 and over.