We conducted a cross-sectional survey involving a convenience sample of 300 caregiver-child dyads with children aged from birth to 21 years presenting for care at an urban pediatric primary care clinic in Baltimore, Maryland. The Institutional Review Board of Johns Hopkins University approved this study.

This study was undertaken at a community teaching hospital affiliated with a major academic center. The primary care clinic was staffed by 6 pediatric providers and 15 pediatric residents (3 residents each afternoon for continuity clinic).The pediatric clinic has an annual pediatric volume of approximately 10,000.

To be eligible, pediatric patients accompanied by their caregiver had to present to the clinic between November 2010 and January 2011 during clinic hours of 09:00-17:00. Patients were excluded if they were non–English speaking because we lacked the resources to interview them.

We developed a questionnaire based upon existing literature [4,11,12] that was piloted on a group of 10 caregivers before study initiation to identify ambiguity. Changes were made to clarify wording before study initiation based upon feedback. We used a final 35-item paper-based questionnaire that included multiple selection and 5-point scale questions. Survey domains included (1) demographic information, including caregiver and child age, sex, race, education, annual family income, and insurance type; (2) caregiver email usage patterns; and (3) caregiver attitudes toward email.

A 35-item survey instrument was distributed to 300 consecutive English-speaking caregivers presenting to the clinic. The survey was distributed to the caregiver by a registrar during the check-in process. The survey was either returned or self-administered and collected at the conclusion of the health care encounter. No incentives were offered to complete the questionnaire, which took approximately 10 minutes to complete.

Sample size calculations were based upon the determination of the proportion of email users in this cross-sectional study. It was predicted that approximately 80% of the surveyed population would be email users. We assumed a 95% level of confidence and set our precision at .05 to yield a sample size of 246. Data analysis was performed with the use of Stata version 9.2 (StataCorp LP, College Station, TX, USA). Frequencies and simple means were calculated for each variable, where appropriate. For items utilizing a 5-point scale, we considered both “strongly agree” and “agree” as agreement with that item. Unadjusted and adjusted logistic regression was used to assess associations between demographic variables and binary attitudinal variables. The amount of missing data for study variables of interest was minimal with an average missing data rate per variable of 0.8% and the largest missing data occurring for the insurance variable at 6.4%. In the regression analysis, we used listwise deletion to account for missing data. Results are reported as odds ratios (OR) and adjusted odds ratios (adjusted OR) with 95% confidence intervals (CI). A P value less than .05 was considered significant.

Most caregivers in our urban population have access to email and are interested in communicating with their child’s providers by this method, although only 11% currently communicate with their provider via email. Overall, attitudes toward email were favorable with three-quarters of email users reporting that email would improve communication with their provider and be satisfying. Despite the generally favorable disposition toward email, most caregivers preferred not to receive test or x-ray results by email. This may be related to the finding that 34% of caregivers remain concerned about security issues surrounding email, which suggests that there are content areas that caregivers feel less comfortable discussing via email. Regarding discussing specific pediatric conditions commonly encountered in primary care, there was much greater acceptance of such topics with no notable differences among the types of conditions.

Our study suggests that caregivers with higher education, higher income, and commercial insurance were more likely to use email to communicate. Furthermore, of those caregivers who did report using email, African-American caregivers and those with incomes less than US $30,000 were less likely to have favorable views toward email as a tool to communicate with their health care providers. Although others have suggested that the digital divide is a function of decreased access to email for these groups [9], our data offer further evidence that disparities persist across racial/ethnic and socioeconomic groups even after controlling for access to email and Internet [10]. There may be factors related to the quality of Internet connection or perhaps cultural preferences pertaining to privacy or other factors as yet not determined. Increased emphasis on the meaningful use of electronic health records has led health care systems to develop patient portals that offer access to health information and frequently allow for 2-way secure messaging between patients and providers. However, recent studies are reporting disparities in adult and pediatric patient portal use along racial and socioeconomic lines [13-15]. A recent review on patient-provider email suggests that email has tremendous potential to improve health care communication between patients and providers and should lead to improved satisfaction and quality of care [16]. Yet, the evidence base to assess the effect of email on clinical communication remains limited and of poor quality [3]. Many questions concerning access, acceptance, privacy, and security issues remain unanswered and established national guidelines are currently lacking.

Study limitations include the small sample size and cross-sectional design. However, we surveyed consecutive caregivers and our response rate was high. Other limitations include bias from our convenience sample at a single urban clinic limiting the ability to generalize our results to other populations. Also, we did not make a distinction between personal and professional email usage (or secure patient portals) which could have implications upon caregiver attitudes toward privacy concerns. Lastly, our survey instrument lacks formal testing for reliability and validity.

Further research should continue to closely monitor for exacerbations of existing disparities in pediatrics as the medical community further embraces email and other electronic data communication methods. Text messaging may serve as another alternative communication modality as it too has been shown to be generally accepted by parents [17]. Investigators should help determine which communication modalities are best suited for conveying specific information such as test results, or providing information about medical conditions while taking into consideration the inherent ethical and privacy concerns raised by all forms of communication [18,19].

Caregivers of children in an urban pediatric primary care practice have access to email and would be interested in communicating with health care providers by this method. African-American caregivers and those in lower socioeconomic groups hold less favorable views toward email communication; thus, the use of email may exacerbate existing disparities in health care delivery. Future studies should examine the reasons for these attitudinal differences.