Response: Re: Interventions to Improving Osteoporosis Screening: an Iowa Research Network (IRENE) Study

To the Editor: Dr. Cox¹ was concerned that confounding factors might have invalidated this study.² Certainly it is probable that at baseline, before any intervention, some clinics will be more likely than others to refer patients for bone density testing. To some extent we took that into account by adjusting for patient educational achievement, baseline rates of testing at each clinic, and other factors. Statistical testing also takes into account nonmeasured intrinsic patient and clinic factors which contribute to variation among the clinics. Because there were only 5 clinics in the study, there was a high bar for finding differences among the intervention groups beyond what might be due to intrinsic factors. Our results suggested that clinic variation was influenced by the combined intervention of chart reminder to physician plus mailed education to patients. The large improvement in the clinics receiving this combined intervention was striking compared with no change for the control clinic. Because the clinics that received the combined intervention were doing well at baseline, it might be anticipated that they would have improved little without the intervention or even declined.

This is a small study, and we cannot state definitely that the interventions tested were helpful. However, we tried to interpret the data fairly using available statistical methods.