Does the Use of an Automated Tool for Self-Reporting Mood by Patients With Bipolar Disorder Bias the Collected Data?

Michael Bauer, MD, PhD; Natalie Rasgon, MD, PhD; Paul Grof, MD, PhD; Laszlo Gyulai, MD; Tasha Glenn; Peter C. Whybrow, MD

Disclosures
In This Article

Discussion

The comparison of the demographic characteristics of patients with bipolar disorder who completed the ChronoRecord validation study with those of published longitudinal studies showed that the findings from all studies were very consistent ( Table 1 ). Although statistical analysis of the differences between the samples is not available, visual inspection shows a striking similarity in the age, gender, diagnosis, age of onset, diagnosis of bipolar I or II, marital status, employment status, ethnicity, and number of daily medications and types of medications received. This comparison suggests that any differences between the demographic profile of patients who were able to use the ChronoRecord software for self-reporting and those who used paper-based tools in other long-term studies of bipolar disorder are small. Similar to previous studies, the long-term consequences of bipolar disorder are apparent in the demographic data from the ChronoRecord validation study. Although the patients are well educated, high rates of unemployment, multiple hospitalizations, and a disabled status are not uncommon. The morbidity reported by the patients in all these naturalistic studies of bipolar disorder was also similar. Patients who self-reported their moods on a daily basis using either the automated or paper-based instrument recorded a 3-fold greater amount of time spent depressed than manic: 30.9% of total days depressed and 11.3% manic for those using ChronoRecord[31] as compared with a mean time depressed of 33.2% and a mean time manic of 10.8% using the paper-based NIMH LifeChart.[28] Similarly, symptoms recorded on the monthly retrospective paper form showed that depression occurred far more frequently than mania (63.6% vs. 33.1%).[29] This may reflect a similarity between patients with bipolar disorder who choose to participate in studies at academic research centers.

No difference was found in the self-rating of mood using ChronoRecord software among individuals grouped by severity of illness according to observer ratings on HAMD and YMRS, in relation to any of the demographic characteristics tested. This also suggests that use of the automated system did not create an obvious bias within the collected data. The limitations of this study include the relatively small size and the limited number of participants who were older adults, from an ethnic minority, with observer ratings of severe mania, and who had an excellent outcome.

Computers have permeated all population segments of industrialized countries. The recent computer revolution is unique in that an unprecedented decline in price has led to a much faster rate of incorporation of computer technology in the world's industrialized economies than for any prior technological change.[32] Thus, a tool that requires daily access does not reduce participation to a few demographic strata. In addition, recent studies have shown that people with disabilities view technology more favorably than those without disabilities, primarily because it provides an opportunity to connect to the outside world and to contact others with similar experiences.[24,25]

The finding that demographics did not impact self-reported ChronoRecord mood ratings for patients grouped by severity of illness was consistent with our prior analysis that showed no relation among missing data and the patients' mood or demographic characteristics.[15] Although continued use of an automated self-reporting tool by patients with a severe mood disorder is counterintuitive, another area of research suggests why a home computer may be such an excellent medium for recording psychiatric symptoms. Many studies have reported that the impersonal aspect computer encourages disclosure of behavior that is sensitive or embarrassing[33] and/or unlikely to receive social support.[34] Computerized interviews have been successfully used to detect risky behaviors in patients from a wide range of social and demographic groups, including potential blood donors,[35] adolescents,[36,37,38] substance abusers,[37,39] pregnant Native Americans,[40] and those at high risk for HIV infection[41] or tuberculosis.[42] A meta-analysis of computer versus face-to-face interviews, using 100 measures from 39 studies from 1969 to 1994, showed that computer administration increased disclosure, especially in medical or psychiatric patients.[43]

There is an ongoing need for larger studies of bipolar disorder that include patients across the entire spectrum of illness who experience all phases of the disorder, regardless of comorbid conditions. Naturalistic, longitudinal studies are a critical component of the methodologies proposed to address this problem.[44,45] Such studies offer high external validity; complement the findings of randomized, controlled clinical trials[46]; and for patients with chronic diseases, are especially well suited to determine treatment benefits and risks, long-term outcomes,[47] and prognostic factors.[48] Furthermore, several reviews of the general medical literature have suggested that well-designed observational studies usually produce similar results to randomized, controlled trails.[49,50,51] It is important to take advantage of the benefits that technology can bring to the instruments used in longitudinal studies, including improving data quality, decreasing the amount of missing data, providing immediate feedback for patients and clinicians throughout the course of a study, standardizing data collection across multiple sites, and increasing patient participation over the long term.[10,15] Technology can also eliminate the costs of data entry and combined with high patient acceptance, may help to include a wider variety of patients in studies, including samples other than those from academic centers.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....