Papers

Deprivation and late presentation of glaucoma: case-control study

Scott Fraser, clinical research fellow ^a, Catey Bunce, medical statistician ^a, Richard Wormald, director ^a, Eric Brunner, senior lecturer ^b. 

^a Glaxo Department of Ophthalmic Epidemiology, Institute of Ophthalmology, University College London, London EC1V 2PD, ^b International Centre for Health and Society, Department of Epidemiology and Public Health, University College London, London WC1E 6BT

Correspondence to: R Wormald r.wormald{at}ucl.ac.uk

	Abstract

Top Abstract Introduction Methods Results Discussion References

Objective: To identify socioeconomic risk factors for first presentation advanced glaucomatous visual field loss.
Design: Hospital based case-control study with prospective identification of patients.
Setting: Three hospital eye departments.
Participants: Consecutive patients newly diagnosed with glaucoma (n=220). Cases (late presenters) were those presenting with advanced glaucoma (n=110), controls were those with early glaucoma (n=110).
Results: Median underprivileged area scores were higher among late presenters (29.5; interquartile range 9.0-42.2) than in the control group (21.3; 6.1-37.4) (P=0.035). Late presenters were more likely to be of lower occupational class (odds ratio adjusted for age and referral centre 20.1 (95% confidence interval 2.6 to 155) for group III compared with group I-II and 86.0 (11.0 to 673 for group IV-V compared with group I-II), to have no access to a car (2.2; 1.2 to 4.0), to have left full time education at age 14 or less (7.5; 2.3 to 24.7), and to be tenants rather than owner occupiers (local authority tenants 3.2; 1.7 to 5.8, private tenants 2.1; 0.7 to 5.8). Effects of deprivation were partly accounted for by family history of glaucoma, time since last visit to an optometrist, and lack of an initial diagnosis of glaucoma by an optometrist.
Conclusions: Area and individual level deprivation were both associated with late presentation of glaucoma. Existing evidence shows that late presentation is an important risk factor for subsequent blindness. Deprived groups thus seem to be at greater risk of going blind from glaucoma. Material deprivation may be associated with more aggressive disease as well as later presentation.

	Introduction

Top Abstract Introduction Methods Results Discussion References

Glaucoma is a disease with major importance in public health and accounts for 13% of all new registrations of blindness annually.¹ It affects about 2% of Europeans aged 40 and over and some four times this proportion in African-Americans and African-Caribbeans.² The European and US populations are ageing, and as prevalence of glaucoma is strongly linked with age ^{3 4} the number of people blinded by the disease is set to increase.

End stage glaucoma causes a particularly profound and irreversible visual loss, but population studies show that only half of glaucoma sufferers are diagnosed and treated at any one time.⁵ Late presentation, when visual field loss threatens central vision, is an important risk factor for blindness related to glaucoma.⁶ Research into determinants of presentation with advanced glaucoma is scarce. We therefore designed this case-control study to determine both social and demographic risk factors as well as ocular and biological factors (presented elsewhere⁷). We report on the role of area deprivation and several measures of individual socioeconomic status in the stage of presentation with glaucoma in the hospital eye service. A link between deprivation and advanced glaucomatous visual field loss at first presentation would provide evidence for systematic inequity of access to effective hospital care.⁸

	Methods

Top Abstract Introduction Methods Results Discussion References

We carried out a prospective hospital based case-control study with recruitment at three independent eye departments in England (Moorfields Eye Hospital, London; Sunderland Eye Infirmary, Sunderland; Harold Wood Hospital, Essex) between September 1996 and May 1997. Participants were eligible for study if they were diagnosed with glaucoma according to the case-control criteria described in the box when they were first examined by the ophthalmologist. Those with a previous definite or possible diagnosis of glaucoma or ocular hypertension were not eligible. The optic disc criteria were shown, by a pilot study, to be good indicators of severity of disease.⁹ Intraocular pressure was obtained from the standard Goldmann tonometer reading at initial examination. We calculated that we needed 110 cases and 110 controls to provide 79% power to detect a threefold increase in the odds of late presentation in a factor present among 10% of the control group at the 5% level of significance (two tailed test).

Patients were excluded if they had problems performing the visual field test. This was defined as having more than one third fixation losses or one third false positive or one third false negative responses for glaucoma on visual field analysis.

We conducted the study prospectively to reduce selection and recall bias. Recruitment after the first of two visual field tests gave a consecutive series of cases and controls. All fields were examined by one author (SF) to ensure consistency of case definition. After patients gave their informed consent they were telephoned by a trained interviewer masked to the case-control status. The interviewer validated demographic data and asked a series of standard questions regarding socioeconomic status (occupational class, car access, and housing tenure), education (age at leaving full time education), ethnic origin (white European, African/African-Caribbean, and Asian), use of general medical services, and use of sight testing (optometric) services. There were no losses to telephone follow up. The study was approved by the ethics committee of Moorfields Eye Hospital.

Referral
The referral of most glaucoma patients to the hospital eye service in the United Kingdom is initiated by optometrists, usually after abnormal findings on sight tests. At the time of our study children under 16 years, people with diabetes, people on income support, and those with a first degree family history of glaucoma and over 40 years of age were exempt from sight test charges; optometrists claimed a fixed fee from the local health authority for these tests. Since this study, exemption has been extended to all people over 60 years of age.

Deprivation measures
We used Jarman's underprivileged area score to classify deprivation.¹⁰ The score is based on weightings derived from general practitioners of the effects of eight census variables on workload in primary care. The variables (general practitioner ratings in brackets) are ward percentages of households with: elderly living alone (6.62), one parent families (3.01), children under 5 years (4.64), social class V (3.74), unemployment (3.34), overcrowded households (2.88), residents who had moved house within the year (2.68), and residents born in the Commonwealth (2.50). The underprivileged area score is calculated by linking individual postcodes with census data.

Statistical analyses
The effect of each socioeconomic factor on the likelihood of presenting with advanced glaucomatous field damage was estimated with unconditional logistic regression.¹¹ We previously showed that late presentation was associated with referral type (not referred by an optometrist with a diagnosis of possible or probable glaucoma), family history (protective), and time since last visit to an optometrist.⁷ These factors were treated as covariates in this analysis. All analyses were conducted with STATA.¹³

	Results

Top Abstract Introduction Methods Results Discussion References

Table 1 summarises characteristics related to demography and glaucoma in the study sample (previously reported⁷)and the socioeconomic characteristics by case-control status. The Jarman score is presented as median (interquartile range); the higher the score, the greater the area deprivation. The largest single occupational group was III (skilled and non-skilled manual). Just over half had access to a car. Most participants had left full time education at the age of 14. Half were owner occupiers, and 41% were in council housing.

Table 2 shows the odds (adjusted for age and centre) of late presentation for each deprivation measure before and after adjustments. Lower socioeconomic status and education level was linked with late presentation, though for underprivileged area score this relation was weak. Adjustment for centre in the age adjusted models had a negligible effect on the odds ratio (data not shown). Precision of the risk estimates for occupational status is poor because there was only one late presenter in group I-II. Ethnicity, added in model 2, accounted for about a quarter of the log odds of late presentation associated with education level and council tenancy. Model 3 also adjusts for cognitive and behavioural factors⁷ to try to show the possible effect of aggressive disease. Compared with the results of model 1, there was a substantial residual effect for occupational group while the effects of educational level and housing tenure were reduced.

	Discussion

Top Abstract Introduction Methods Results Discussion References

We found consistent evidence for an association between lower socioeconomic status and late presentation with glaucoma. Patients who presented with more advanced field loss had higher underprivileged area scores, lower occupational status, and lower education level and were less likely to have access to a car and more likely to be tenants. As in cancer, presentation with advanced glaucoma is associated with a poor prognosis. ^{6 14 15}

The inverse association between socioeconomic status and late presentation can be interpreted in different ways. Firstly, socially patterned differences in health seeking behaviour are likely to operate. This was so for the reported use of optometry services in the general household survey (1991-4).¹⁶ Regular sight testing was associated with higher social class and in our present study greatly reduced the risk of late presentation.

Alternatively, long term deprivation may lead to more rapidly progressive and aggressive disease. The links between raised cortisol concentration, ocular hypertension, and glaucoma¹⁷ provide some support for a psychosocial mechanism mediated by altered hypothalamic-pituitary-adrenal function.¹⁸ Thus the possible "length bias," which could be an important determinant of both case and control status, might be driven by pathological mechanisms linked to social status.

How might socioeconomic factors, including educational deprivation, influence risk of late presentation? Leaving school by the age of 14 was no longer a significant predictor (model 3, table 2) after adjustment for recall of family history of glaucoma and time since last visit to the optometrist. Both factors are likely be influenced by level of education, which influences awareness of the disease and the need for regular sight testing. While the effects of education and occupation are correlated and therefore difficult to separate,¹⁹ it is interesting to note that the risk factor associated with social status remained highly significant in the same model. This may suggest additional mechanisms such as a diet insufficient in micronutrients²⁰ and lifetime stress,¹⁸ which add to the aggression of glaucoma in poorer people.

In addition to occupational status previously published results from this study⁷ indicated that intraocular pressure at presentation, family history of glaucoma, method of referral to hospital, and the number of years since the last visit to an optometrist were found to be independently associated with late presentation. The data strongly suggested an association between intraocular pressure and advanced field loss at presentation, with a 1.2 (1.12 to 1.28) increase in the odds of late presentation per unit increase in mm Hg after adjustment for the other mentioned factors.

Those with a family history of glaucoma were found to be about one third (0.29; 0.12 to 1.28) as likely to present with advanced field loss than those with no family history. People referred from any source, other than an optometrist with the correct diagnosis (of glaucoma), were estimated to be four and a half times more likely to be late attenders (adjusted odds ratio 4.53; 1.52 to 13.48). The data also provided strong evidence that the longer since the last visit to an optometrist the greater the likelihood of first presenting with advanced glaucomatous visual field loss (adjusted odds ratio per year 1.25; 1.10 to 1.42).

We also found a strong association between African-Caribbean ethnic origin and late presentation.⁷ The Baltimore Eye Study found African Americans to be at significantly increased risk of visual impairment.²¹ In our study sample ethnic origin accounted for some, but not all of the association between socioeconomic status and late presentation (table 2).

Our study adds to the sparse evidence that lower socioeconomic status is linked with increased risk of chronic eye disease²² and extends it to include glaucoma. It has been shown previously that late presentation with amblyopia in childhood is linked with deprivation²³ and that adult urban Americans of lower social status have higher rates of visual impairment.²¹ Lower socioeconomic class is also associated with poor uptake of mammography and cervical screening,²⁴ and social deprivation has been linked to later presentation of cancers including breast,²⁵ colorectal,²⁶ and skin.²⁷

Implications for health care
Early detection of glaucoma is clearly desirable, but the means to achieve this on a population basis remains problematic. Not only is there a lack of a single adequate screening tool but it is not clear how the test can be delivered to those most in need.²⁸ Optometrists are encouraged by their council to perform diagnostic testing for glaucoma on all their patients aged over 40 who present for routine sight testing. This places a strain on both optometrists and the hospital eye service as a result of false positive referrals.²⁹ A further major problem, the lack of good evidence for the effectiveness of lowering intraocular pressure, is being tackled in a Swedish trial (early manifest glaucoma trial, www.nei.nih.gov/neitrials).

Conclusions
This hospital based case-control study is appropriate because it analyses factors for late presentation to hospital. A weakness of our study is that it examines access only to the NHS. There was only one case from occupational group I-II, which might indicate that those from higher social groups with advanced disease may seek private health care. We cannot know whether selection bias led to an underestimate or overestimation of the effect of individual social deprivation. The fact that in the control group there were 33 participants from the higher socioeconomic groups suggests that this source of bias is unlikely to be serious.

	Acknowledgments

We thank Peter Phelan (Sunderland Eye Infirmary), Rod Daniel (Moorfields Eye Hospital), and Charles Claoué (North East London Eye Partnership) for their enormous help in recruitment of the patients.

Contributors: All the authors contributed to the design of this study and wrote the paper jointly. RW and SF initiated the research. SF, RW, and EB designed the patient questionnaires. SF was responsible for coordinating the study and recruiting the patients. CB provided the data analysis. RW is the guarantor.

	Footnotes

Funding: International Glaucoma Association and Moorfields Locally Organised Research Scheme. EB is supported by the British Heart Foundation.

Competing interests: None declared.

	References

Top Abstract Introduction Methods Results Discussion References

1.	Evans J, Rooney C, Ashwood F, Dattani N, Wormald R. Blindness and partial sight in England and Wales: April 1990-March 1991. Health Trends 1996; 28: 5-12.
2.	Sommer A, Tielsch JM, Katz J, Quigley HA, Gottsch JD, Javitt J, et al. Relationship between intraocular pressure and primary open angle glaucoma among white and black Americans. Arch Ophthalmol 1991; 109: 1090-1095[Abstract].
3.	Tielsch JM, Sommer A, Katz J, Quigley HA, Royall RM, Javitt J. Racial variations in the prevalence of primary open angle glaucoma. JAMA 1991; 266: 369-374[Abstract].
4.	Klein BEK, Klein R, Sponsel WE, Franke T, Cantor LB, Martone J, et al. Prevalence of glaucoma. Ophthalmology 1992; 99: 1499-1504[Medline].
5.	Tielsch JM. The epidemiology of primary open angle glaucoma. Ophthalmol Clin North Am 1991; 4: 649-657.
6.	Grant WM, Burke JF. Why do some people go blind from glaucoma? Ophthalmology 1982; 89: 991-998[Medline].
7.	Fraser SG, Bunce CV, Wormald RPL. Risk factors for late attendance of chronic glaucoma. Invest Ophthalmol Vis Sci 1999; 40: 2251-2257[Abstract/Free Full Text].
8.	Smeeth L, Heath I. Tackling health inequalities in primary care. BMJ 1999; 318: 1020-1021[Free Full Text].
9.	Fraser SG, Bunce CV, Wormald RPL. Retrospective analysis of risk factors for late presentation of chronic glaucoma. Br J Ophthalmol 1999; 83: 24-28[Abstract/Free Full Text].
10.	Jarman B. Identification of underprivileged areas. BMJ 1983; 286: 1705-1709.
11.	Sloggett A, Joshi H. Higher mortality in deprived areas: community or personal disadvantage? BMJ 1994; 309: 1470-1474[Abstract/Free Full Text].
12.	Office of Population Censuses and Surveys. Standard occupational classification. London: HMSO, 1991.
13.	Stata Corporation. Stata statistical software release 5.0. College Station, TX: StataCorp, 1997.
14.	Wilson R, Walker A, Dueker DK, Pitts-Crick R. Risk factors for rate of progression of glaucomatous visual field loss. Arch Ophthalmol 1982; 100: 737[Abstract].
15.	Mikelburg FS, Schulzer M, Drance SM, Lau W. The rate of progression of scotomas in glaucoma. Am J Ophthalmol 1986; 101: 1-6[Medline].
16.	General household survey: analysis of ophthalmic data 1990-91 to 1993-94. London: Government Statistical Service, July, 1995.
17.	Schwartz B, McCarty G, Rosner B. Increased plasma free cortisol in ocular hypertension and open angle glaucoma. Arch Ophthalmol 1987; 105: 1060-1065[Abstract].
18.	Brunner EJ, Marmot MG. Social organisation, stress and health. In: Marmot MG, Wilkinson RG, eds. Social determinants of health. Oxford: Oxford University Press, 1999.
19.	Davey Smith G, Phillips AN. Confounding in epidemiological studies: why "independent" effects may not be all they seem. BMJ 1992; 305: 757-759.
20.	Stallone DD, Brunner EJ, Marmot MG, Bingham SA. Dietary assessment in Whitehall II: the influence of reporting bias on apparent socioeconomic variation in nutrient intakes. Eur J Clin Nutr 1997; 51: 815-825[CrossRef][Medline].
21.	Tielsch JM, Sommer A, Katz J, Quigley H, Ezrine S. Socioeconomic status and visual impairment among urban Americans. Arch Ophthalmol 1991; 109: 638-641.
22.	Quigley H, West S, Rodriguez J, Munoz B, Klein R, Snyder R. Prevalence of glaucoma in a population-based study of Hispanics. ARVO abstract 2896;March 2000.
23.	Smith LK, Thompson JR, Woodruff G, Hiscox F. Social deprivation and age at presentation in amblyopia. J Public Health Med 1994; 16: 348-351[Abstract/Free Full Text].
24.	Calle EE, Flanders WD, Thun MJ, Martin LM. Demographic predictors of mammography and Pap smear screening in US women. Am J Public Health 1993; 83: 53-60[Abstract/Free Full Text].
25.	Wells BL, Horm JW. Stage at diagnosis in breast cancer: race and socioeconomic factors. Am J Public Health 1992; 82: 1383-1385[Abstract/Free Full Text].
26.	Powe BD. Fatalism amongst elderly African-Americans. Effects on colorectal cancer screening. Cancer Nursing 1995; 18: 385-392[Medline].
27.	Robinson JK, Altman JS, Rademaker AW. Socioeconomic status and attitudes of 51 patients with giant basal and squamous cell carcinoma and paired controls. Arch Dermatol 1995; 131: 428-431[Abstract].
28.	Wormald RPL, Rauf A. Glaucoma screening. J Med Screen 1995; 2: 109-114[Medline].
29.	Griffiths PG. A surfeit of screening? BMJ 1997; 315: 318[Free Full Text].
30.	Report of the independent inquiry into inequalities in health. London: Stationery Office, 1998 (Acheson report).

(Accepted 14 December 2000)