Background: High response rates to patient surveys are important for ensuring that results are representative of the views of their local population. In the English NHS Patient Survey Programme, areas with high ethnic diversity and younger populations have lower response rates, with London-based organisations having the lowest response rates in England. One reason commonly cited for this reduced response rate is the higher proportion of patients in the areas for whom English is not their rst language.

Objective: To investigate if o ering the option to complete the inpatient questionnaire online in English or a number of other commonly spoken languages helps to increase response rates, particularly for hard-to-reach groups.

Methods: This online pilot survey was run alongside the national 2008 Inpatient Survey. Five London hospital organisations each generated an additional sample of 500 recent inpatients. Identical questionnaires were mailed to all patients but the additional sample were given the opportunity to complete the questionnaire online, in English or one of the ten most commonly spoken non-English languages in London. These participants were also able to call a free telephone number to request a paper copy of the translated questionnaire (in one of the ten non-English languages).

We compared the response rates and response patterns of the additional sample to the 2008 National Inpatient Survey sample to investigate the e ect of providing an online completion option in the native language of the participants and uptake of telephone translation services.

Results: In four of the ve participating organisations there was no di erence between overall response rates for the online completion option and national survey samples. Generally, uptake of the online option was low in all languages (1% of sample), especially so for non-English completion (0.2%). Cost per response was unacceptably high for this research. Telephone completion of the questionnaire, in up to 127 languages, has traditionally been off ered in the National Inpatients Survey. Uptake of this option in a non-English language halved from 2007 to 2008, with no di erence between the online and national samples.

Conclusions: O ering an online completion option did not increase response rates to the survey overall, nor did providing translated questionnaires speci cally improve the response rates for those from non-White ethnic groups. There are issues other than the language spoken which result in decreased response rates for those from non-White ethnic groups.