Sir,

We would like to thank Dr Muula for his comments regarding our recent publication.1

The purpose of this paper was not to precisely define the prevalence of disability in a population, but to identify a vision crisis that has not previously been recognized and to put a best estimate on the prevalence of this problem in a very marginalized community.

The first issue Dr Muula raises concerns how representative our study sample was of the general inner-city population in Vancouver's downtown eastside (VDES). We agree that our data may underrepresent individuals who do not attend medical care and have noted this in the second-last paragraph of the paper. However, we must emphasize that our subjects were not attending the Vancouver Native Health Society (VNHS) for eye examinations as Dr Muula suggests. Instead, these individuals were there for general, nonophthalmic care (paragraph 2 of the Methods section). As a result, we believe, there is no selection bias towards eye disease in our sample. Moreover, the dates and times of each intake clinic were varied over the course of the 2-year study period, and were not conducted at the same time of day or on the same day of the week. As such, we believe that we achieved as representative a sample of clinic attendees as possible. We also know that demographic data from the VNHS clinic has been found to correspond quite closely to the larger VDES community.

Dr Muula also has concerns regarding the use of a single ophthalmologist for the eye examinations in our study. We do not believe this is a valid criticism. First, our study did not require specific patient diagnoses, only a simple categorization of the aetiology of vision loss—a routine practice for ophthalmologists. Second, although it would have been interesting to have more than one physician confirm our ocular classification, such an approach was not practical from a physician availability standpoint and would not necessarily have improved our categorizations. Third, contrary to Dr Muula's comments, all of the ophthalmic diagnoses in our prior study of a medium-sized Canadian city (Prince George) were also made by a single ophthalmologist.2 This latter study was a chart review and, as such, the patients' ophthalmologists were occasionally consulted if there was diagnostic uncertainty for the physician performing the data abstraction.

We agree that there are methodological differences between our VDES and Prince George studies. These differences were unavoidable given the dissimilarities of the medical and social environments in these communities. Our intergroup comparisons are not intended to be unqualified; however, the prevalence figures for our VDES population (even taking into account the lower bound of the confidence intervals) do suggest that this is a distinctly different cohort of individuals with alarmingly high rates of vision loss (up to 10 times higher) compared with the general Canadian population.