Objectives: (1) To pilot, and evaluate, a tonometry training intervention for general practitioners; (2) to evaluate the efficacy of three types of tonometer (Perkins, Schiotz and Tonopen) in the hands of these general practitioners; (3) to evaluate the predictive value of tonometry in screening for glaucoma in a general practice population; and (4) to evaluate the acceptability of tonometry to general practice patients. Design: After being trained, three general practitioners recruited 73 of their patients over 50 years of age to attend for measurement of intraocular pressure by tonometry. Intraocular pressure was initially measured by an ophthalmologist using the Goldmann applanation tonometer, and then recorded in random order by GPs using three types of tonometer - the Perkins, Schiotz and Tonopen. Setting: A group general practice in a middle-class suburb in southern Brisbane. Outcome measures: (1) Comparison of measurements on the Perkins, Schiotz and Tonopen tonometers with that of the 'gold standard', the Goldmann Applanation tonometer; (2) prevalence of disc and perimetric abnormality suggestive of glaucoma among those patients with increased intraocular pressure; and (3) the acceptability of contact tonometry to general practice patients. Results. There was considerable variability between intraocular values obtained across doctors and across instruments (6% to 95% of values within 4 mm of the 'gold standard'). The Schiotz tonometer provided the most uniform assessment of intraocular pressure across the groups. Nineteen of the 73 patients (26%) had intraocular pressures greater than or equal to 21 mmHg using the 'gold standard'. Of these, 18 were followed up with funduscopy, gonioscopy, repeat measurement of intraocular pressure and visual field assessment. Five had persistent elevations of intraocular pressure greater than or equal to 21 mmHg on subsequent assessment, and two had mild abnormalities of cup-disc ratio with normal fields on testing with Humphrey computerised perimetry. Conclusion: No one hand-held tonometer proved highly accurate in the hands of all three doctors. Even measured optimally, increased intraocular pressure alone was a poor predictor of glaucoma. Of the population screened, two patients (3%) showed evidence of mild cup-disc abnormality requiring follow-up.
