Recovery and maintenance of vision following treatment in patients with exudative age-related macular degeneration (AMD) are highly dependent on the stage at which patients present to the clinic for evaluation and treatment.1^–7 Unfortunately, many seek care when marked visual deterioration has occurred,3 4 6 7 even to levels of legal blindness,3 6 7 limiting the outcome and cost-effectiveness of the treatment. The reasons why these patients present late are not well established.

Socio-economic deprivation has been associated with presentation at advanced stages of the disease in breast cancer,8 cutaneous melanoma9 and glaucoma,10 among others. Various indices of deprivation have been developed to study the socio-economic status of individuals, including the Jarman, Townsend and Carstairs scores11^–13 and the Scottish Index of Multiple Deprivation (SIMD).14 The latter has several advantages. It is a weighted combination of six domains: current income, employment, housing, health, education, skills and training and geographic access and telecommunications;14 across these there are 31 indicators of deprivation (Jarman, Townsend and Carstairs are based on 10, four and four indicators, respectively). Thus, the SIMD represents a very complete method of assessing socio-economic deprivation. The area of analysis used by the SIMD is the data zone (on average 750 residents), which is substantially smaller than the wards or postcode sectors used in other deprivation indices, allowing for the identification of small “pockets” of deprivation and being more homogeneous with regards to its constituent population and household socio-economic characteristics.

The purpose of this study was to investigate the influence of socio-economic factors on visual acuity (VA) at presentation in patients with exudative AMD. The effect of other parameters on presenting vision was also studied.

SUBJECTS AND METHODS

The medical records of all consecutive patients with newly diagnosed exudative AMD examined at the Grampian University Hospitals—NHS Trust, Aberdeen and Gartnavel General Hospital, Glasgow, between 1 July 2004 and 30 June 2005, were retrospectively reviewed. There were 240 patients (240 eyes), 155 (64.5%) females and 85 (35.4%) males, with a median age of 79 years (interquartile range (IQR), 73–83 years).

Demographics, duration of symptoms, VA, exudative AMD characteristics including type (predominantly classic, minimally classic, occult or disciform scar) and location (extrafoveal, juxtafoveal or subfoveal) of the choroidal neovascularisation (CNV), status of fellow eye and patient’s address were recorded. VA was obtained using Early Treatment Diabetic Retinopathy Study (ETDRS) or Snellen visual acuity charts, with the patient’s best current refraction. VA was converted into decimal values for the statistical analysis of the data.

The SIMD 2004 was used to determine the socio-economic status of the patients. For this purpose, the postcode of each patient was recorded and used to determine the data zone to which the patient belonged.14 The SMID scores and ranks for that particular datazone were obtained from a spreadsheet downloaded from the SIMD 2004 website.14 This spreadsheet contains the scores of the various indicators of deprivation evaluated by SIMD, including an overall SIMD score, which was used in this study. Each data zone has a rank, from 1 to 6505, based on its SIMD score; the lower the rank, the more deprived the area. Patients belonging to data zones ranked between 1–1301 fall within the category of “most deprived” as they form the lowest 20% of the population.

The effect of age, SIMD score, VA in fellow eye, location and type of CNV and whether the patient was living in the area (local authorities) served by the hospital at Glasgow or Aberdeen on VA in the study eye at presentation was investigated using general linear modelling. The analysis was undertaken (1) considering SIMD score and VA as binary variables (lowest 20% or upper 80% for SIMD scores and ⩾0.5 or <0.5 for VA) and (2) considering SIMD scores and VA as continuous variables. Analyses were carried out using SPSS Version 14.0 (SPSS, Chicago).

The median duration of symptoms in different groups of patients was compared using Mann–Whitney U tests. Comparisons of the proportions of patients with different attributes were carried out using χ² tests. Comparisons with regards to VA among the different types and locations of CNVs were undertaken using general linear modelling.

RESULTS

Demographis, duration of symptoms, visual acuity, CNV characteristics and distribution of SIMD scores have been summarised in table 1.

The duration of symptoms prior to presentation between patients from most deprived (median duration 4 months, IQR 2 to 6 months) and least deprived areas (median duration 3 months, IQR 1.5 to 6 months) was not statistically significantly different (p = 0.137, Mann–Whitney U test). The duration of symptoms was not different either (p = 0.167, Mann–Whitney U test) between patients with good VA in the contralateral eye (VA⩾0.5) (median duration 3 months, IQR 2 to 6 months) and those with lower levels of vision (VA<0.5) (median 3 months, IQR 1 to 6 months).

Only type and location of the CNV in the affected eye, and age of the patient had a statistically significant effect on VA at presentation (p<0.001, p<0.001 and p = 0.003, respectively). The age of the patient was negatively associated with VA at presentation, with older people presenting with lower levels of vision. SIMD scores, area (Glasgow/Aberdeen) and VA at presentation in the fellow eye were not significantly associated with presenting vision (p = 0.959, p = 0.247 and p = 0.056, respectively). The lack of an effect of SIMD scores was consistent in both areas, as there was no significant interaction between SIMD score and area (p = 0.902).

The above findings remained unchanged independently of whether VA and SIMD scores were considered binary or continuous variables, when tertiary referrals from outside the area (Local Authority) covered by each hospital were removed from the analysis (n = 75), or when patients with VA tested using ETDRS charts (n = 54) were removed from the analysis (and only those with Snellen vision remained).

Overall no evidence of an effect of SIMD on VA at presentation was found. The 95% confidence interval for the difference in median VA between those from the poorest 20% of datazones and those from the richest 80% was 0.07 to −0.06, suggesting that if there were to be an effect of deprivation on VA at presentation, this would be very small. Comparing only people from the richest 20% of datazones with those from the poorest 20% of datazones (n = 56 and n = 44 respectively) there was also no significant difference in median VA at presentation (p = 0.892).

DISCUSSION

Socio-economic deprivation has been associated with late presentation and subsequent poor vision and blindness in eye conditions such as glaucoma10 and amblyopia.15 In exudative AMD, however, the association between lower socio-economic status and poor vision at presentation was not found herein.

Chew and colleagues16 evaluated clinical features and socio-economic factors in 115 patients with subfoveal CNVs <3000 μm in basal diameter undergoing either self funded photodynamic therapy (PDT) or government-funded transpupillary thermotherapy (TTT). All patients were offered PDT, and those that declined it received TTT. Patients not willing to pay for PDT and who received TTT had worse macular disease (larger lesion size and poorer vision) and lower average annual income than those receiving PDT; no difference in the levels of education between groups was found. The authors concluded that the severity of exudative AMD appeared to be associated with lower socio-economic status. The discrepancy between findings in the above and the current study may be explained by methodological differences between the two. The current study included a larger group of consecutive patients (n = 240) with newly diagnosed exudative AMD, independently of whether they received treatment, and the SIMD 2004, which includes more indicators of deprivation than the annual income or the level of education, was used to evaluate the socio-economic status of the patients.

The fact that no statistically significant differences on VA in the study eye at presentation were found between patients with good and those with poor vision in the fellow eye is surprising and worrisome. It would be expected that patients with poor VA in the fellow eye should note new symptoms in their only remaining “functional” eye and seek immediate care. Furthermore, it is likely that many of these patients would have had a previous evaluation at the time the vision in the first eye had failed, and appropriate counselling with regards to returning urgently at the onset of symptoms. This finding suggests that frequent follow-up of high-risk patients and/or the use of sensitive, home-based, self-administered visual function tests by these patients may be needed in order to warrant early diagnosis and treatment.

Although the results presented herein are limited by the retrospective nature of the study, this research indicates that all patients with neovascular AMD are equally vulnerable to present to clinics with low levels of vision independently of their socio-economic status and suggests that educational programmes aimed to improve the awareness of this disease and its early detection and treatment should be extended to patients of all socio-economic levels but particularly to the oldest group. Screening strategies to detect early disease are needed to warrant early diagnosis and preservation of vision, especially now that new effective treatments are available.