Background Policy architects are keen to emphasise that releasing information on the quality of hospitals into the public domain will drive patients away from poor quality providers and thus motivate system wide improvements. However, previous evidence on the impact of releasing quality information to the public have been mixed, with various reasons being proposed as possible barriers, including patients being unaware of the information available, being unable to understand it, or feeling it to be irrelevant. We examined three cases of high profile reports into failings in quality and safety at English NHS hospitals in order to assess whether this negative publicity was associated with a decrease in patient numbers. These reports all garnered significant media attention and presented a stark picture of performance and so we hypothesise that such negative coverage may have induced a fall in patient numbers.

Methods We analysed trends in admission for nonemergency admissions with a difference-in-difference approach, using four separate comparison groups: (1) matched on previous hospital numbers; (2) the five closest hospitals geographically; (3) hospitals in the same administrative area; (4) a national comparison group. We also employed waiting time as an instrumental variable to account for possible effects on waiting times on patient numbers. We used three outcomes: overnight admissions; day-case admissions; and patients not attending for their booked appointments, and used a model segmented into three month periods after report publication in order to examine the range of short and longer term effects.

Results Two out of three of the hospitals we examined showed no difference in patient numbers at any point in the 12 months after negative reports were published. One of the hospitals showed a reduction in overnight patients for 3 months (–12%, p=0.003) and 6 months (–14%, p<0.001) after report publication, but patient levels returned to normal after this. Additional examination of patient numbers in advance of these reports coming out also did not change these results.

Conclusion The publication and dissemination of highly critical reports by a health care regulator does not appear to have resulted in patients’ sustained avoidance of the hospitals involved. We only found an impact in one of the three hospitals we examined, and this only persisted for 6 months. This reminds us that simplistic assumptions regarding the power of information to drive patient numbers are unrealistic, and that reliance on these alone to drive quality improvement may be misguided.