The Institute of Medicine’s vision for a high quality healthcare system includes the concept of patient-centeredness, which emphasizes the need to be responsive to patient preferences and needs [1]. Towards this goal, the use of email between physicians and patients is recognized as enhancing communication [2-4] and is generally favored by many patients [4,5]. Moreover, with the public’s increased use of the Internet, the use of email between physicians and patients will likely increase. Yet, physicians’ adoption of email to communicate with patients has been relatively slow with only modest increases in adoption rates in recent years [6,7].

The current literature on the subject of physician-patient email is generally focused on somewhat limited populations or attributes. Work has been done, for example, on the experiences of early physician [2,8] or patient [9] email users, physician attitudes [10] and concerns [11] towards using this communication medium and the general benefits [2,3,12] of doctor-patient email. Early work has identified appropriate content for physician-patient email and has highlighted the medico-legal issues associated with this practice [13-15]. Published work has also examined the nature and regularity of email inquires by patients [14] or their caregivers [16] with physicians [5]. Early studies have reported relatively low rates, generally between 6 [17] and 10 [5] percent, for physician-patient email. However, these previous statistics have typically been reported from surveys of patients and not necessarily of physician groups or practices. As a result, despite the increasing attention in the literature, few recent scholarly studies have comprehensively examined the frequency of physician-patient email use or the factors associated with this practice.

To help interested doctors benefit from email communication with patients, the American Medical Association (AMA) and the American Medical Informatics Association (AMIA) have adopted sets of guidelines for physicians [18,19]. It is unknown to what extent physicians comply with these best practice recommendations while emailing patients. The current paper specifically examines these issues directly by scientifically surveying a large sample of physicians in the state of Florida. In addition, it identifies numerous trends in mid-2005 that update previously identified developments in the use of physician-patient email.

A total of 4203 returned surveys were available for the current study. This represents a 28.2% participation rate. Demographic and practice characteristics of the respondents are shown in Table 1. Overall, demographics of respondents did not differ from known characteristics of Florida physicians [20].

Patient-provider electronic mail has been previously defined as “computer-based communication between clinicians and patients within a contractual relationship in which the healthcare provider has taken on an explicit measure of responsibility for the client’s care” [19]. As such, it is an important tool for physician communication with patients in both general [21,22] and specialized [23,24] areas of medical practice. Despite the improved communication potential from the use of physician-patient email, the number of physicians electing to do so is still low, even though broadband Internet access is very common. Our finding of over 85 percent of physicians having high-speed Internet access is consistent with other US-based surveys [25].

Yet, the present study, conducted in mid-2005, found that only 16.6% of physicians in Florida used email with patients, and only 2.9% of the overall respondents used it frequently. This latter number, derived from physicians’ responses, suggests how rare email communication remains in clinical practice and is substantiated by studies showing the low number of patients who have ever sent email to a physician [5,26]. Although some patients do not yet have regular access to email [11], studies of the general public show both an increasing access to email accounts [27] and a general interest in email communication with their physicians [5,12]. From the perspective of the diffusion theory, physician-patient email is only now beginning to traverse the uphill slope of the adoption curve [28]. Yet, the fact that physicians are regularly using email from their offices to communicate with virtually all other entities (except patients), indicates that barriers seem to be specifically impeding email use with patients.

These barriers have been identified previously [2,15,29] and appear to be due to several specific fiscal and legal causes. Even though most email communications are asynchronous in nature, physicians spend valuable time and resources responding to email messages from patients [8,30,31]. This represents an “opportunity cost” to some physicians, particularly if the email system in place does not replace other modes of communication such as telephone messages, postal letters, etc [2]. In addition, the purchase and maintenance of encryption software, required to achieve maximum privacy, adds expense to the practice [32]. Only recently have several pilot programs in the United States begun to reimburse physicians for the expenses associated with direct email consultation [33-35].

The pace of email communication to patients has also been slowed by concerns from physicians [30] and staff [36] over general liability and privacy stemming from the recent Health Information Portability and Accountability Act [37-39]. For the interested reader, several excellent reviews exist that discuss the numerous legal and policy implications of physician-patient email and electronic health record use [40-43]. For those interested in the policy issues related to unsolicited email from patients, a seminal study by Eysenbach and Diepgen, which describes the policy implications, is recommended reading [44].

There may be a difference in perceptions between patients and physicians of the benefits accrued from the use of electronically available information. For example, a survey of patient use of the Internet for health information suggested that patients perceive more benefits and fewer risks than their physicians do, when this mode of information gathering is utilized [45]. A further study of these perceived differences in email benefits/risks is warranted.

Another important observation from the current study is that the use of email with patients occurs most frequently among certain groups of physicians. In one of the few studies that reported demographic information of physicians who do, and do not, regularly email patients, Gaster et al found that female physicians, younger physicians, and university-based clinic physicians were proportionately more likely to use patient email [10]. Community-based physicians, who more often offer primary care, tended to use email less than university- or county-hospital based clinics. In the present study, both family medicine physicians and surgical specialists were more likely to email patients. We believe the percentage of surgical specialists using email may be higher because they tend to work in larger practices (which were also more likely to use email). Family medicine doctors also have a higher likelihood of email use in Florida. We hypothesize this may be due to an ongoing health information technology educational program actively being pursued by the Florida and American Academies of Family Physicians, respectively. Similar to findings by Gaster et al, we found less email communication by older physicians. We believe this trend will disappear as the current physician workforce ages and younger physicians, with a higher general comfort level with information technology, appear in the workforce.

As email communication differs from traditional, written medical communication between physicians and patients and among providers, guidelines for best practices have been developed. These guidelines have emanated from both the medical [19,46-48] and health informatics [18] professions, as well as experts in the bioethics [49,50] and legal [51,52] fields. In the current study, we chose to design our survey questions around the guidelines found in two large US medical and informatics organizations because of their breadth and general availability [18,19]. The AMIA released its guidelines in 1998, and the AMA [18,19], in 2000. Both of these sets of recommendations are available online for physicians to review and utilize.

One of the most important findings of the current study is that few physicians were routinely utilizing these guidelines for email communication with patients, despite their broad availability for several years. In this regard, the current study results are similar to those of Gaster and colleagues from a 2000-2001 survey of physician practices related to email use [10]. They found that 75% of physician-respondents never or rarely obtained consent to communicate with patients by email, 66% never or rarely discussed confidentiality or security concerns and 58% never or rarely documented email in the patient record. Importantly, a separate study by White et al found that the majority of patients involved with regular physician email communication do follow guidelines when they are educated about their nature and importance [14]. The findings by White et al, done from the patient’s perspective, coupled with the physician-oriented findings from our current study, suggest to us that the main barriers to guideline use may be more with the physician’s initiation than with the patient’s compliance.

The low rate of adherence to published physician-patient email guidelines may have several reasons. Among these reasons may be the lack of knowledge about the existence of guidelines by many practicing physicians; the lack of agreement with the guidelines (eg, not feeling that the guidelines are required in their particular practice), or an impracticality to their implementation. Unfortunately, the present study was not designed to determine reasons for not adhering to these recommended guidelines. However, given the results presented in the current study, the medical profession should consider further educating physicians about email communication, assess the barriers facing implementation, and better understand the practicality of utilizing the guidelines themselves.

We acknowledge that there are several important limitations of this study. First, we recognize that the survey response rate, although higher than comparable previous studies [22,53,54], may be a limitation. However, upon employing common methodologies used to detect bias, we failed to identify the presence of response bias. Second, as with other self-reported surveys, the study relies on the willingness and ability of participants to give accurate responses. Finally, because the purpose of the study was to identify the use of email by physicians in one state, the results of this study should be generalized to other geographic regions with caution.

To enhance email communication between physicians and patients, we believe that further work to educate both physicians and patients on the advantages and limitations of email correspondence is necessary. In addition, efforts are needed to deal with the fiscal barriers many physicians face in the regular use of email as a quality-enhancing tool in patient care. Although we are encouraged by recent efforts to reimburse physicians for email communication in several areas of the United States, most US physicians do not yet have access to these reimbursement programs. As these barriers are addressed in the United States, we believe email communication between physicians and patients will become better defined, better compensated and a resource for better clinical care of patients.