Shum M, Kelsh MA, Sheppard AR, Zhao K. An evaluation of self-reported mobile phone use compared to billing records among a group of engineers and scientists. Bioelectromagnetics. Sep 20, 2010. Ahead of print.
Exposure assessment in retrospective case-control studies of potential adverse health effects of radiofrequency radiation is primarily based on self-reported mobile phone use. Accuracy of these self-reports has always been a concern.
The objective of this study was “to examine the accuracy of self-reported usage of mobile phones compared to billing record information.”
Study participants were 60 employees of a scientific consulting company who had used a company-provided mobile phone for at least one year. Mobile phone use information from questionnaires and detailed billing records were examined over a longer period (3 years) compared to other studies. Information on both incoming and outgoing calls was available from billing records. Recall accuracy was assessed by age, gender, self-selected reporting period (daily, weekly or monthly), and time period.
Based on data from questionnaires, duration of calls was shorter than that based on billing records (average duration 2.1 min vs. 2.8 min; p<0.01). Average number of calls per day reported by participants was higher than that based on billing records (7.4 vs. 4.6; p=0.002). Spearman correlation coefficients were 0.71 and 0.69 for average duration of calls and average number of calls per day, respectively. Up to 50% of the participants could not recall the actual number of calls or the duration of calls to within 50%. Recall accuracy was roughly comparable for men and women, stable during the 3-year period, slightly better for older participants (≥40 years), and much better for those who reported on a monthly basis than for those who reported on a weekly or a daily basis.
Interpretation and Conclusion
These results are consistent with findings from other studies showing considerable variability in recall and modest correlation between self-reported mobile phone use and data from billing records. Unlike other studies, no difference in recall accuracy over time and no tendency for a decrease in accuracy with age were observed. Limitations of this study include a small sample size and a low participation rate (~30%). Generalizability of these findings may be limited because recall in this group of professionals who use mobile phones primarily for business purposes may be different from recall in other population groups. Since the study was not designed to examine differences in recall between cases and controls, its impact on epidemiological results is unclear.