Objectives Social deprivation as measured by the Index of Multiple Deprivation (IMD) and Townsend scores has been shown to be associated with advanced presentation of primary open angle glaucoma. The aim of this study was to investigate the putative association of social deprivation as a risk factor for acute primary angle closure (APAC) in a UK urban population.
Methods Case notes of 139 consecutive patients presenting with APAC at the Birmingham and Midland Eye Centre, Birmingham, UK, were examined. Deprivation was scored using the IMD 2004 and Townsend scores. These score were compared with the West Midlands reference population.
Results The level of deprivation in patients with APAC was graded according to the IMD quintiles, in which quintile 1 represents the highest level of deprivation and quintile 5 represents the lowest level of deprivation. Of the patients studied, 66.1% (n=90) were from quintiles 1 or 2 (most deprived) whereas 9% (n=12) came from quintile 5 (least deprived), compared with predicted frequencies of 40% and 20%, respectively. Deprivation levels measured by frequency within each IMD quintile were significantly higher in the APAC group compared with the reference population (χ2, p<0.001).
Conclusions In this population patients presenting with APAC were more likely to come from areas with a high level of social deprivation.
- IMD score
- social deprivation
- ReGAE, Research into Glaucoma And Ethnicity, Townsend score
Statistics from Altmetric.com
- IMD score
- social deprivation
- ReGAE, Research into Glaucoma And Ethnicity, Townsend score
The link between social deprivation and health is well established for a number of specific conditions such as hypertension and heart disease, and for overall morbidity and life expectancy.1–5 Acute primary angle closure (APAC) is a known sight-threatening ophthalmic emergency that is a significant risk factor for preventable glaucoma-related blindness, and is part of the spectrum of primary angle closure glaucoma (PACG)—a major health-burden across south Asia.6–8 The following risk factors have been identified in patients presenting with APAC: shallow anterior chamber, family history of angle closure, age >40 years, female sex and Asian ethnicity.9 Although increased social deprivation has been shown to be associated with late presentation of primary open angle glaucoma, its association with PACG has not been explored.10
A well-validated measure of social deprivation is the Index of Multiple Deprivation 2004 (IMD 2004).11 The IMD incorporates seven domains of deprivation: income, employment, health deprivation and disability, education skills and training, barriers to housing and services, crime, and living environment. Each super output area (the unit of geographical area) is given a score and a ranking for each domain. Thus an estimate of a person's level of deprivation may be deduced from their postcode.
We also utilised the Townsend score, which is an alternative well-validated measure of social deprivation.12 Prior to the introduction of the IMD score it was generally regarded as the gold-standard for measuring material deprivation. The Townsend score includes four variables: unemployment, overcrowding, lack of owner-occupied accommodation and lack of car ownership. The score is a summation of the standardised scores for each variable (scores greater than zero indicate greater levels of material deprivation).
The aim of this study was to investigate the relationship between the level of social deprivation and APAC in a UK urban population.
Patients in this study were retrospectively identified from detailed searches of: (1) Accident and Emergency records/database at the Birmingham and Midland Eye Centre (BMEC), Birmingham, UK; (2) the laser safety records at BMEC; and (3) the inpatient database at BMEC, for a period of 60 months. Within the Greater Birmingham catchment area, the agreed ophthalmic policy is to admit all patients with APAC for emergency management. The BMEC is the only hospital in this area with a 24-h dedicated ophthalmic casualty unit and ophthalmic inpatients beds.
A diagnosis of APAC was made according to validated standardised criteria13 14 Patients presented with at least two of the following symptoms: ocular or periocular pain, nausea and/or vomiting, or a previous history of intermittent blurring of vision with haloes. On examination, presenting intraocular pressure has to be be more than 21 mmHg plus at least three of the following signs: conjunctival injection, corneal epithelial oedema, mid-dilated unreactive pupil, shallow anterior chamber; together with the presence of an occluded (iridotrabecular contact) angle in the affected eye, verified by gonioscopy.
Patients with secondary angle closure associated with neovascular glaucoma, or uveitis were excluded.
Demographic data (age, sex, ethnicity, postcode) were documented and the level of social deprivation was measured according to the Townsend score and the IMD 2004 as indicated by the patient's postcode. The distribution of IMD scores for our study group were compared with the distribution for the catchment area (the West Midlands) using the following quintiles: quintile 1 represents the highest level of social deprivation ranging to quintile 5 with the lowest level of social deprivation. The observed and expected frequencies for these quintiles were compared using a χ2 goodness of fit test. We also performed a series of generalised linear model analysis allowing for ethnicity, sex and age.
At the electoral ward level, comparison of APAC patient IMD scores with local ward IMD mean was achieved with paired t tests on the difference between patient IMD and ward IMD mean.
One-hundred and thirty-nine subjects were deemed eligible for inclusion in the study. Of these, 96 were women and 43 men, and the mean (SD) age was 68.2±11.6 (range 31–96) years. Eleven subjects had both eyes affected at presentation (eight women, three men) so only one eye was randomly chosen and included for the purpose of this study. Ethnicity coding demonstrated that 110 patients were white, 19 were of Indian Asian ethnicity, four were African-Caribbean and three patients could not be coded. Three patients (two women, one man) had no complete records of their postcodes and were therefore excluded from the analysis.
The deprivation score of the remaining 136 participants (as judged by their individual postcode) were found not to be significantly different from the ward that they came from, overall (p=0.91) and individually for each ward, suggesting both that they were typical of their ward and that there were no significant fluctuations within the wards considered.
On comparing the APAC patient frequency in each West Midland IMD quintile with the expected frequency of 20% using a χ2 goodness of fit test, we found that there was a statistically significantly higher proportion of patients with high levels of deprivation (66.1% in quintiles 1 and 2) than predicted from the reference population (40%) (p<0.001). Conversely we found a much lower proportion of study patients in the least deprived group (quintile 5) than predicted (table 1). Comparing deprivation by means of the alternative Townsend deprivation score also showed a significant association of APAC with the level of deprivation (p<0.001) (table 2).
In this study we have shown that patients presenting with APAC have higher than predicted levels of social deprivation as measured by the IMD 2004 and Townsend scores.
Fraser et al have previously demonstrated that late (delayed) presentation of glaucoma with advanced (potentially blinding) glaucomatous optic neuropathy was significantly associated with higher levels of social deprivation. It should be noted that their population was composed of chronic glaucoma cases; therefore the majority of patients had primary open angle glaucoma, although cases of chronic PACG were not excluded. The study of Fraser et al pre-dates the IMD index: they employed the earlier Jarman's underprivileged area score. Like the IMD and Townsend scores, the Jarman score is based on postcodes, but is a less complete assessment of deprivation.15 Fraser et al did, however, also include a number of individual surrogate markers for deprivation including occupational class, housing tenure and access to a car.10
One of the limitations of studies that use an area-based index of deprivation such as the IMD is that they measure the score of a geographical area rather than an individual. It can be argued that those living in a deprived ward are not necessarily deprived individuals. However, such significant local fluctuations are found to be uncommon, and indeed in this study we have demonstrated that the patients (at least as judged by their full postcode) are typical of the wards from which they came. It should also be noted that as a global index, incorporating seven separate domains, incongruities in the overall IMD score are sometimes observed. In such cases it is worth examining the individual domain rankings and scores to find out which is primarily contributing to a low IMD. This is not only useful to guide regeneration strategies (ie focusing on particular needs to be targeted), but it may help health professionals understand unexpected clusters of health problems.
In this study, our reference population for social deprivation was the entire West Midlands region. We acknowledge that the West Midlands contains more wards of increased social deprivation than other UK populations. We addressed the possibility that the majority of referrals to BMEC (a supra-regional tertiary referral unit) may be from a smaller catchment area, which might hypothetically have had higher levels of social deprivation than our standard reference population. However, an average IMD score for Birmingham, Walsall, Dudley, Sandwell and Solihull (the immediate catchment area) was also significantly different from the IMD of the angle closure group (32.77 vs 36.0; p=0.038). Although the absolute measurement of APAC incidence was not the purpose of this study, trends of incidence can be compared since the denominator for calculating incidence is the population size in each quintile (which will be equal). Comparison of these frequencies with the expected frequencies therefore examines a trend of incidence with level of deprivation.
It is important to consider the reasons why deprivation may be associated with APAC. APAC is a condition in which the pathogenesis is primarily anatomical in nature, and might thus be assumed to be more embryologically influenced than environmentally influenced. In APAC the primary problem centres on an episode of acute pupil block occurring within an anatomically small anterior segment/eye in association with a disproportionately large (often cataractous) lens.
Socially patterned differentiation in health-seeking behaviour does exist. This was reported for the use of optometry services in the general household survey (1991–1994),16 and regular sight testing was associated with higher social class and reduced risk of late presentation.10
We have considered possible reasons for the association we have found between this “anatomical” disease and social deprivation. It is known that communities with high levels of social deprivation have poor access to optometric and hospital-based eye-care services. There are many potential barriers to accessing primary eye-care services. Recent work has demonstrated that perception of optometric credibility can affect health-seeking behaviour. In particular, some individuals perceive that there is a potential conflict of interest in the dual role of optometrists as providers of primary eye-care screening and as commercial outlets for the selling of glasses.17
The net effect of this barrier to accessing the primary eye-care services is that patients from areas of greater social deprivation are less likely to have ophthalmic problems diagnosed at an early stage. Conversely, in areas of the population with low indices of social deprivation, health-seeking behaviour is more pro-active, costs are less likely to be a deterrent, and the pro-dromal stage of APAC can be diagnosed well before the acute crisis occurs; in addition, patients with anatomically narrow drainage angles can receive prophylactic laser peripheral iridotomies.
The enlarging, ageing lens is a major factor in the pathogenesis of APAC. It is known that there is an increased prevalence of cataract in areas with high levels of social deprivation, and this might be explained by higher levels of smoking, poor diet, and increased prevalence of associated medical conditions such as diabetes mellitus.18 In view of the fact that individuals from areas of high social deprivation present late with all ophthalmic problems, it is likely that these communities have increased numbers of patients with more advanced cataract (increased lens thickness). This will predispose these individuals to an increased risk of APAC.
Hyperopic refraction is considered to be a risk factor for PACG.14 Hyperopia is strongly associated with short axial length, small anterior segment and small horizontal corneal diameter. There is emerging evidence of an association between hyperopia and low levels of educational attainment. A study in Wales investigated the relationship between hyperopia and education test results in a cohort of primary school children, and concluded that there was evidence for a link between hyperopia and impaired literacy standards in children.19 These findings theoretically could contribute to the lower social status in later life of this group of patients with short axial length eyes. We acknowledge that the level of educational attainment is an important factor in disease-awareness behaviour and in early health-seeking behaviour.10
This is the first study to our knowledge to investigate the role of social deprivation in APAC, suggesting that there is a significant association between APAC and higher levels of social deprivation (as measured by two independent indices) in a large multi-ethnic urban population in the UK.
There is an increasingly strong body of evidence linking blinding glaucoma and greater levels of social deprivation. Eye-care and other health professionals need to make sure that appropriate resources are directed to our most vulnerable populations.
Part of this data was presented at the American Academy of Ophthalmology Annual Meeting 2008 as a poster.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.