Price Survey of Assistive Devices and Supports for Persons with Disabilities - December 2003

1. Content Development

The list of products started with a base of supports included in the 1991 Health and Activity Limitations Survey (HALS) and those listed in the Income Tax Act regulations for the Medical Expense Tax Credit. From this base point, the Applied Research Branch worked with the Roeher Institute to enhance the draft conceptual framework of disability—related products, supports and services. From that point, consultations were undertaken with members of other disability-related non-governmental organizations (for example, the Canadian National Institute for the Blind (CNIB) and the Canadian Hearing Association) and members of the community to ensure that the framework was reasonably all encompassing and accurate (See Appendix A for a list of the organizations consulted). However, the sample is meant to be representative and not fully comprehensive of all aids/devices and supports available. The sample¹ of goods and services and average prices provided in this report is provided in Appendix B.

Prices reported reflect market prices before subsidization by national/provincial/territorial governments and/or insurance plans.

In addition, distributor, province and city data were collected. Where multiple prices were obtained from a single distributor, a distributor identifier was added.

2. Universe, Sample Frame

The Applied Research Branch worked with the Prices Division of Statistics Canada to carry out a national price survey of commercial retailers, service providers and other supports to obtain prices of goods and services.

For most of the data collected, the sample frame was created from which commercial retailers and service providers were selected for the survey from the business register, the Electronic Yellow Pages, and the Canadian Register of Psychologists. However, there were a number of exceptions.

• Supports for the blind where the CNIB was the sole source of price data. CNIB also provided sales data, and price estimates for many of these supports were developed using sales based weights;
• Drugs used in the treatment of diabetes and psychiatric conditions, results were taken from the November 2001, Consumer Price Survey.

In most cases, sampling was limited to cities currently included in the Consumer Price Index.²

Survey results were undertaken through the fall of 2001 and winter of 2001-02. Where possible, the prices obtained were as at October, 2001.

3. Response Rates and Sample Sizes

• Price data on disability supports was obtained from a total of 307 respondents of 851 surveyed (36%).
• A survey of disability related service fees and the costs of orthoses/prostheses were distributed to 940. Of these, 540 (58.3%) responded with fees/prices.
• Prices of drugs from 99 suppliers and prescription glasses/lenses were obtained from 114 suppliers. Both were taken from the November, 2001 Consumer Price Survey.
• Public transportation prices were obtained from 58 municipal transportation authorities and taxi fees from 57 local providers.
• Prices for van conversions were obtained from 17 respondents (44.7%) of 38 providers solicited.

4. Reporting

Chapter III provides summary results of the price survey. It also includes brief descriptions of each of the devices priced. In most cases, it then provides average prices at the national level. Where there is sufficient sample size to assure reliability, average prices are also reported for sample brands.

Nation wide average prices are always reported where there were 25 or more prices available for a product. The price is provided for smaller samples of a product (a minimum of 10) when tests using descriptive of location/dispersion statistics (coefficient of variation, skewness, kurtosis, etc.) indicate that the mean is likely a "good" point estimate of sample prices. In addition, individual Coefficients of Variation (CV) (ratios of the estimated standard error of the mean to the mean) were calculated and some data is either not reported or reported with a caution using the following criteria:

1. Where less than 16.5% are reported without an asterisk.
2. Those with a CV which lies between 16.6% and 33.3% are reported with an asterisk. The asterisk addresses caution when using these prices because the average price reported has a high standard error.
3. Where CV was greater than 33.3%, the price data was not reported.

Appendix C provides details regarding the tests undertaken and the criteria used for the decisions made.

The use of these inclusion criteria allowed the reporting of prices for support devices and aids for all disability types except assistive devices for the speaking impaired.

In two instances, where sample sizes were under 25, single price observations were excluded in the calculation of mean prices because they were "extreme outliers" and had inordinate effects on this statistic. In each instance, the price excluded was more than three times higher than the next highest price reported.

Where there is adequate sample size (at least 15 observations and Appendix C criteria are met) prices are also provided for various brands/models of a device. Where appropriate or where there are insufficient sample sizes of brands, ranges of brand/model averages are provided.

Chapter IV reports on the effect of three factors on the prices of a subset of the aids, devices and services sampled:

• Region;³
• Size of community (as characterized by the logarithm of the area's population);
• Distributor pricing pattern (where prices for multiple goods were provided by the same distributors).

Analysis was undertaken for classes of devices where there is sufficient sample size for each device (an average of 15 observations per region and at least 5 observations in each region) to establish clear trends using analysis of variance/covariance and each retailer (at least 4 prices) using factor analysis.⁴ Differences are reported from the national mean and, where relevant, between regions in Chapter IV where, with 95% confidence, it could be inferred that average prices were either higher or lower.

In Chapter IV regional average prices are provided, where applicable. Where community size likely had a significant effect on price, the averages reported have been adjusted to remove the estimated effect of regional differences community sizes.

• ¹This "judgmental" sample was developed because there was insufficient market information concerning the sales distribution of specified goods and services to allow for a probabilistic sample.
• ²Some goods and services related data was collected from other, small communities, primarily located in Atlantic Canada.
• ³There was consistently inadequate sample size to allow inferences for some provinces/territories. As a result, suppliers in the Atlantic Provinces were grouped as were Manitoba and Saskatchewan. Suppliers in the territories were grouped with adjacent provinces unless there were large variations in Territorial prices vis-à-vis the rest of the country. Thus, Yukon observations were usually merged with the British Columbia samples and Northwest Territories/Nunavut prices with those of Alberta.
• ⁴Specifically, residuals of the prices from individual product analyses of regional/provincial differences for each class of aids/devices/services were factor analysed when outlets provided four or more prices. In each analyse, a single major factor emerged. The proportion of variance characterized by this factor was attributed to outlet-specific price variation.