Kakkar A, Jacobson BC. Failure of an Internet-Based Health Care Intervention for Colonoscopy Preparation: A Caveat for Investigators. JAMA Intern Med. 2013;173(14):1374-1376. doi:10.1001/jamainternmed.2013.6477
Copyright 2013 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Internet-based tools for health care delivery are proliferating. We examined the effectiveness of an online instructional video aimed at improving bowel preparation prior to colonoscopy. We hypothesized that an educational tool explaining the importance of preparation, the precolonoscopy diet, and how to administer the bowel purgative would lead to improved bowel preparation. However, what we learned during this randomized clinical trial of 2000 patients was that there may still be considerable limitations to reliance on the Internet for health care interventions.
Consecutive outpatients booked for colonoscopy by open-access at Boston Medical Center were randomized for this institutional review board–approved study. The video, available in English and Spanish languages, was posted on our hospital’s website. In addition to standard instructions, patients in the intervention arm were also mailed a 5 × 7-in card with instructions in English and Spanish stating, “Please visit www.bmc.org/digestivedisorders to watch an instructional video about preparing for your colonoscopy.” This card was stapled to the front upper left corner of the written instructions. A nurse included a reminder to watch the video during our standard precolonoscopy telephone call.
We collected information about sex, age, race, ethnicity, education, insurance, Internet access, how many times (if any) they watched the video, and aspects related to bowel preparation. We used median income published by the Internal Revenue Service for patients’ zip codes as a surrogate for income. We measured the association between demographic variables and Internet access using logistic regression. Trend testing was performed with the Cochran-Armitage trend test.
After exclusions for missing bowel preparation information and failure to attend colonoscopy, we had 387 patients in the intervention arm and 350 controls. There were no significant demographic differences between the groups (data not shown). Among the 387 patients who received instructions to view the online video, only 24 (6%) reported watching the video, despite written and verbal instructions to do so. There were no technical problems reported by those instructed to watch the video, although we did not inquire as to why they did not view the video.
There were no differences in bowel preparation related end points. However, among 595 patients who provided information about Internet access, we found that only 378 patients (64%) had access (279 with Internet access at home; 99 with access only outside the home). We found no independent associations between having Internet access and sex, race, ethnicity, income, or type of insurance. However, age, and level of education were strongly associated with Internet access (Table). Older patients and less-well-educated patients were significantly less likely to have internet access (P value for trends, <.001). Age (P = .99) and education (P = .25) were not associated with likelihood of viewing the video.
We are unaware of previous trials examining the effect of Internet-based patient education on bowel preparation for colonoscopy. While a majority of our patients had Internet access, only a tiny fraction followed instructions to watch a brief, online video to help prepare for screening colonoscopy. Moreover, we identified older patients and those with less education as 2 populations at risk for lacking Internet access.
While studying Internet-based efforts to alter clinical outcomes has become increasingly popular,1- 5 like us, others have found limited impact on patient compliance or failure to improve the efficiency of health care delivery.1,5 Furthermore, previous investigators have reported that particular racial and ethnic groups and those with low income are less likely to use the Internet for health education or personal health records.6,7 However, these studies failed to account for education, which we found to be an important variable to explain differences in Internet access.
Our study was conducted at a single urban academic medical center, which may limit generalizability. Nonetheless, our patients represent a diverse mix of race, income, education level, and insurance status, suggesting that our findings may apply to other settings.
Our findings have important implications for future interventions aimed at improving health outcomes through the Internet. First, patient-directed educational endeavors that require additional efforts by patients, such as watching online videos, may be limited by low participation rates even in the setting of Internet access. Second, such endeavors should include assessments of why patients fail to use online resources to allow appropriate alterations in strategy. Third, more aggressive efforts to help subjects view the video might have increased usage rates. For example, our nurse could have walked subjects through the process of accessing the video on the reminder call. Fourth, Internet-based patient resources may not be suitable for all older patients or those with limited education. This becomes an important consideration as the nation prepares itself for online insurance exchanges and an expansion of Medicare’s Physician Compare website.
The Internet provides a unique platform for disseminating health information. However, limited patient interest and lower rates of internet access among older and less well-educated patients should be considered when planning Internet-based health care initiatives.
Corresponding Author: Brian C. Jacobson, MD, MPH, Section of Gastroenterology, Boston University Medical Center, 85 E Concord St, Room 7721, Boston, MA 02118 (email@example.com).
Published Online: June 3, 2013. doi:10.1001/jamainternmed.2013.6477
Author Contributions:Study concept and design: Kakkar, Jacobson.
Acquisition of data: Kakkar, Jacobson.
Analysis and interpretation of data: Jacobson.
Drafting of the manuscript: Kakkar, Jacobson.
Critical revision of the manuscript for important intellectual content: Jacobson.
Statistical analysis: Jacobson.
Administrative, technical, and material support: Kakkar, Jacobson.
Study supervision: Jacobson.
Conflict of Interest Disclosures: None reported.
Previous Presentation: Findings from this study were presented as a poster at Digestive Disease Week (Kakkar A, Jacobson BJ. Internet-based education and “physician compare” websites may discriminate against older and disadvantaged patient populations); May 7-10, 2011; Chicago, Illinois.
Trial Registration: clinicaltrials.gov Identifier:NCT01099553