Automated telephone outreach with speech recognition did not improve rates of colorectal cancer screening

Automated telephone outreach with speech recognition failed to improve colorectal cancer screening rates, according to the results of a randomized controlled trial.

Researchers selected 80,000 men and women aged 50 to 64 years from the Harvard Pilgrim Health Care plan. Half were randomly assigned to automated calls and half to usual care. The health plan serves roughly 1 million people in Maine, Massachusetts and New Hampshire. Of the participants, 10,432 in the intervention group and 10,506 in the usual care group had not been previously screened and were eligible for the analysis.

The automated system engaged participants in a conversation about screening; obtained whether or not the member had ever been screened for colorectal cancer; offered information about methods of screening; and described the potential for screening to reduce disease burden.

The researchers said that one year after the calls, the screening rate for those in the intervention group was 30.6% compared with 30.4% in the usual care group (P=.76). There was no observed intervention effect after adjusting for covariates (OR=1.01; 95% CI, 0.94-1.07).

In multivariate analysis, there was a small effect of the intervention on the likelihood a participant would undergo colonoscopy (OR=1.08; 95% CI, 1.00-1.16). Among those in the intervention group, 21.4% had a colonoscopy compared with 20.3% in the usual care group (P=.04).

Survival analysis showed no statistically significant difference between groups for time to any screening, but time to colonoscopy was shorter for those in the intervention group (P=.03).

Simon SR. Arch Intern Med. 2010;170:264-270.