The Internet and the digital environment have contributed to widespread feelings of information overload. People in professional occupations feel the imperative to keep up with their professional development, and new research is being made available faster than ever before because of the World Wide Web. Primary care physicians, that is, physicians who are general practitioners with office practices who are not often attached to a specific hospital, other medical institution, or academic institution have specific information needs; among them, the need for professional development. How can a primary care physician keep up? The primary care physician is at a disadvantage compared to the hospital doctor. Doctors affiliated with hospitals or academic medical facilities have access to extensive health sciences libraries and the librarians within. They often have easier access to diagnostic and prescriptive digital aids such as personal digital assistants. They can access the licensed databases held by the libraries, and they also have access to many colleagues who can provide anecdotal information based on their years of medical experience. The primary care physician may have some of these informational tools, but generally not all, and possibly none. This paper explores the information needs of primary care physicians, and their information-seeking and information retrieval behaviour, and outlines how digital reference service can best be utilized to meet the needs of these physicians. With training, an affiliation to a health sciences network, and support, these physicians can get their information needs met, and can keep up in a virtual sea of medical information.
Are the Questions Being Answered?
The literature concludes that many questions generated by patient visits in the primary care setting are not being answered. It is noted that more than "50% of patient visits" to a general practitioner not aligned with an academic institution "generated a patient care question; only 33% of these were pursued and answered" (Graber, Bergus, and York 520). Another study concludes that on average, one question is generated for every 10 patients seen. Of those, 57% are pursued, and answers are found around 70% of the time (Gorman, "Info. Needs in Primary Care," abstract). The study done by Ely et al. recorded "3.2 questions for every 10 patients seen" (318). The numbers may vary, but the evidence is overwhelming that for a variety of reasons, primary care physicians are not pursuing answers to all their questions.
Primary care physicians in a general practice have several options when it comes to retrieving information. In a survey of 300 Ohio physicians (185 responded, and 47% of those were primary care physicians), Sprague and DeMuro discovered that "personal print collections, conferences, colleagues, and continuing medical education" were listed by nearly all of the respondents as ways they sought information (28). Gorman found much the same results, with a focus on human contact for answering questions (733). He points out the use of narratives by physicians to communicate clinical questions: many of the physicians he interviewed would place the clinical question in the larger context of the patient's background to increase understanding of the question. Gorman equates the reliance on stories to communicate the information need with heavy reliance on human sources for answers to clinical questions. It appears as though the need to contextualize goes hand in hand with the desire to find answers to patient care questions. This focus on human contact will play a role in designing digital reference services for primary care physicians, which will be explored later in the paper. Sprague and DeMuro also cite pharmaceutical or medical device company reps as sources of professional information, mentioning that previous studies have reported that "drug information is the most frequently sought information" by physicians (30). And the two researchers also list a variety of library resources used: books or journals, the librarians, and literature searches (30).
The studies consistently report that many doctors choose not to pursue patient-care questions. However, it is heartening to know that when they do choose to seek answers, they are successful 80% of the time (Gorman, "Excellent Information"). The situation begs the question of why the physicians are not hunting down answers. Time constraint is a big reason. Busy general practitioners do not have unlimited time to search out every question evoked by patient care. Because of the increase in information and "the complexities of modern health care practices, clinicians could spend hours to weeks reading texts and seeking expert opinions for each patient they encounter" (Westberg and Miller 7). One study found that physicians spend an average of less that 2 minutes looking for the answer to a question, while that study and others report that MEDLINE searches by medical librarians can take upwards to 30 minutes (Ely et al. 318). Physicians need information that is current, authoritative, accurate, and quick. The study by Ely et al., which quantifies the questions asked by physicians relating to patient care, discovered that "[b]usy family physicians need bottom-line answers to their questions, and they need them quickly" and that evidence is "most useful when it has been digested into quickly accessible summaries" (318). The Cochrane Health Library would be a good primary source for physicians because it consists of systematic reviews of prior practice and research (see Appendix A for the Cochrane Health Library and other online sources). With the presence of the Internet, and the ability for physicians to hook up online with a larger network of doctors and health institutes, the answer to the patient-care and medical knowledge questions dilemma is digital reference.
The Issue of Access
When thinking about primary care physicians utilizing libraries for their medical information needs, the issue of access on many levels comes to the fore. Many general practitioners do not have access to health sciences libraries attached to academic institutions or hospitals, and thus do not have direct access to library-held databases, collections, or the information professionals at the institution. A quick scan of the Edmonton Public Library databases reveals only two possible databases that physicians could use for authoritative information: the Cochrane Health Library, specializing in clinical resources and thus appropriate for physicians, and the Health and Wellness Resource Center, which seems to slant more towards the patients as users, rather than the doctors (Edmonton Public Library website). And there are even more physicians in small towns and rural setting who do not have access to even the most rudimentary library. The issue of access moves away from libraries as physical spaces and towards digital environments. Not every primary care physician is online, and Gorman concludes that "rural physicians reported less access to some information resources" than urban physicians ("Infor. Needs in Primary Care," abstract). It is not unlikely that one of those resources is the Internet. Thus, these doctors do not have access to web-based databases, search engines, online journals, or other online medical resources. Access also involves the ability to use the resources available. Doctors can have access to all the relevant sources and databases, but if they are not trained in the choice of database, effective searching techniques, and recognizing authoritative sources, the technology is useless. Training of some kind, combined with user-friendly software, is necessary to get the most out of digital resources.
Funds, Training, and Time: Information at One's Fingertips
Provided that the technology is available, a personal computer and an Internet connection in the primary care physician's office would be a good source of digital reference material. There are several considerations when thinking of an implementation like this. Cost is a factor. If the doctor shares an office with one or more physicians, they could share the cost of the hardware, software, Internet access, and the fees involved in subscribing to certain online databases and journals. A solo doctor would have to absorb the cost alone. While it is conceivable that a person can get a PC, hook up to the Internet, and find all he or she is looking for, a physician would do well to consider some form of training. The doctor would have to know the basics of the computer hardware in order to make an informed choice when purchasing it, and to have a basic idea of how it all fits together in the office. Software training is necessary as well, so the physicians can decide what software will be right for their individual needs. Once the computer is up and running, education about databases (online and CD-ROM), the Internet (databases, email, listservs, forums, newsgroups, mailing lists, medical portals, etc.), and library catalogues (local availability) is the next step.
Because of time constraints, training must be considered in order to make the most effective and efficient use of time taken to seek out answers to questions. As end-users tend to overestimate their own searching abilities, the physician must learn effective search methods, including natural language and controlled vocabularies, as well as keyword searches and Boolean operators. For self-searches by physicians, controlled vocabulary would be useful provided they have access to an information retrieval thesaurus or subject headings. Most databases have an online copy of their controlled vocabularies. However, if the medical term involved is new, there will be a lag time in it being picked up by the controlled vocabulary; thus, instruction in free text searching is necessary as well. There also must be training as to how to recognize what search techniques are used on the various databases and in the various catalogues. Education is also important in terms of being able to formulate the searchable and answerable question. Hersh and Hickman report that "with minimal training, physicians could improve their performance to the level of experienced searchers by their fourth online search" (1350). This study involves the use of the MEDLINE database. And although it might be a stretch to conclude that with minimal training, someone can become an experienced searcher in only four searches, the study does indicate that training is beneficial. Physicians must be made aware of the different types of searching that they can do on the Internet. For example, Graber, Bergus, and York found that search engines and portals (or "meta-lists") are two distinct entities. Using a medical portal "requires medical knowledge and an intuition that is not needed with a search engine. To find an answer with a meta-list, the searcher has to make appropriate choices" (523). The doctors must also be trained to think in terms of bias and authority, especially in regard to Internet sources. Evaluating the authority of the source is necessary in order to find the best information. The doctors need to be aware of who sponsors particular websites and who distributes e-reference services and databases, and what the implications of those associations are.
Ideally, primary care physicians will have the funds, training, and time to set up a personal digital reference service and do their own searching, and thus meet their information needs in patient care and continuing medical education. Indeed, many doctors are already online. A 2001 Harris Interactive survey of 400 physicians revealed that "89 percent of the respondents said they used the Internet, of whom 9- percent said they used it to find clinical information" (Blumenthal 525). However, studies also show that the questions generated by patient visits still are not being pursued. Have things really changed in the information revolution? Westberg and Miller list several relatively current barriers to physicians seeking information:
Limited access to resources
Cost of resources
Difficulty learning or using resources
Poor organization of resources
Variable quality of information (7)
In 1985, Covell, Uman, and Manning studies self-reported information needs of 47 physicians. During the patient visit, "only 30% of physicians' information needs were met," often by "another physician or another health professional" (596). Back in the pre-online 80s, the researchers anticipate the perfect set-up:
The technological dreams of nearly 20 years ago have become a reality; yet the same statistics are being generated in terms of information needs that arise from patient care, and how few of those needs are being met. In discussing barriers to having desired information sources in the office, Covell, Uman, and Manning, in 1985, note that "13% [of the physician respondents] cited a glut of information sources of differing reliability as a barrier to finding needed information" (598). If this was a concern in 1985, imagine how overwhelmed the physicians are now. I believe that the reason the statistics for pursuing answers to questions are the same now as the have been in the past is because the glut of information sources has become a torrent, and the means by which to find the answers has become far too complex. It is likely that doctors fresh out of medical school will find the information torrent easier to navigate. Indeed, Mittman and Cain, in a report forecasting the future of the Internet in health care, predict that "[u]sage of the Web for enhancing clinical knowledge will increase as newly minted doctors who have been trained to do online search come into private practice" (31). But because online searching is such a new event, older doctors may simply be overwhelmed, and thus immobilized, but the information revolution. It takes time and money in order to keep up to date with the latest technological advances and the information that is constantly being pushed via these new means.
Print material that can be accessed via a desktop computer terminal [. . .]
is still over the horizon. A system that incorporates the efficient organization
of appropriate, up-to-date textbooks and periodicals with easy retrieval of specific
points of information needed in office practice would be a boon to today's
If primary care physicians are reluctant to set up their own digital reference programs, other options are potentially possible. An affiliation with a hospital library or an academic health sciences library would provide general practitioners with databases, online journals, and support in the form of the librarians. If the physicians are in the same city, they could utilize the institution in person or use the digital reference service. If the situation involves rural or small town physicians, the digital reference service would be the main line of action. Information could be accessed via a service such as e-reference (reference services by email or web forms and templates) or chat-reference (a real-time reference encounter using Internet chat software). There are some considerations involved with such an arrangement. There would be increased use of the resources found at the medical library facility. The institution would have to decide if the librarians would be available to help the physicians at any time, or if there would be only certain hours in which the doctors could access the digital services. The added traffic of the general practitioners utilizing the resources would affect the licensing agreements for e-journals and databases. Policy would have to be created which addresses the type of services offered, and their set hours, as well as the cost (if any) to the primary care physicians for their use of e-journals and databases. Provided that these issues could be worked out, an affiliation with an medical institution or health sciences library would help the physicians with patient care questions. In terms of continuing medical education, the institutions could be pro-active in setting up distance education opportunities using the digital environment.
Westberg and Miller envision a similar arrangement, and they have created a model for health sciences institutions to allow for ease of integration of the additional users. Their model "focuses on the modern academic health centers as the most logical site to integrate and distribute a wide variety of both electronic and human information resources for primary care" (12). They propose a "triage system"-Web-based, and mediated by the Internet (12). The first step would be to train and remotely connect primary care physicians to the academic health centre libraries, essentially giving them remote access to the full range of electronic services offered. Steps two and three would involve software to allow the general practitioners to submit queries they could not find the answers to themselves. Step two would entail deeper searching by the physician via templates or forms, while step three consists of human intervention. After the physician exhausts all his or her options in terms of self-searching, the librarian steps in with help and education. The triage system allows the physicians to do their own autonomous searching. However, if they are not finding the information they need, they can resort to expert help. This model assumes that the doctors would try the search on their own first, and not immediately access the reference librarian. Such an assumption is not assured, as it is well documented that doctors prefer human interaction when answering a clinical question.
Working in the Digital Environment
Although digital reference service can offer quick, authoritative service, one aspect of reference that does change with the electronic environment is the reference interview. Much is lost in the digital environment in terms of body language, facial expressions, and other non-verbal cues found during the traditional reference interview. As well, conversational interaction is stilted and choppy in an email or chat reference context. With email reference questions, unless the doctor and the librarian have ample time for multiple email correspondence, the question must be answered to the best of the librarian's ability based only on the information provided in the email. With chat reference, more interaction is possible, but the non-verbal cues are absent, and the back and forth of the questioning is more like a two-way radio than a telephone in terms of flow. The face-to-face reference interview is an iterative process. This process cannot be fully articulated in the digital environment. Thus, new ways to elicit the information must be used in the online reference interview. Janes and Silverstein differentiate between the synchronous and asynchronous modes of conducting a digital reference transaction. Synchronous (happening at the same time) reference service is chat reference, with the librarian and the physician interacting in real time online. Asynchronous (not happening at the same time) reference service consists of questions being sent in via email or a web form. With asynchronous service "there is more concern about the structure of the inquiry, in the use of forms and guidelines for email messages," while the "people designing synchronous digital reference services are attempting to make them more like traditional services in the use of a real-time interview and response process" (Janes and Silverstein). While on the surface it appears that digital reference interviews might be lacking in terms of not being able to interpret non-verbal cues, Straw comments that there are fewer restrictions when interacting via email. He discusses how librarians and patrons are not bound by the physical, and can interact from anywhere in the world. Another advantage is that because the librarian and the user are not face to face, a situation that can cause pressure for the librarian to find the right answer quickly, the librarian can "step back and think about the range of responses to a user's problem" (Straw). Chat reference is essentially more like the face-to-face reference encounter. However, the conversation is stilted, and because the two participants cannot see one another, the reference interview must be conducted in a different fashion. Librarian Sara Weissman (quoted in DeCandido) claims that the traditional full reference interview is "baggage" from a reference librarian's "desk days" and that the librarian should "give the patrons something, engage them, start the conversation" right at the beginning of the chat reference session.
"HON's mission is to guide lay persons or non-medical users and medical practitioners to useful and reliable online medical and health information. HON provides leadership in setting ethical standards for Web site developers." In an attempt to set standards for online health resources, this site provides a code of conduct (HONcode) consisting of eight criteria for good Web health sources: HONcode
A commercial database that offers a wide selection of information from medical texts, journals, peer-reviewed guidelines, patient handouts and more, all in full text. Graber, Bergus, and York found that this site was the top site for information retrieval, providing answers for the most clinical questions and following all of the Health on the Net criteria for good web-based health information (see above). This is a fee service, but free trials are available.
A portal. This sight claims to "list the best sites that list the sites." A directory of medical directories, Hardin Meta Directory subscribes to the HON code of conduct.
The Canadian federal government's health site. It is updated daily, and contains timely and current information for health care professionals and the general public.
This sight is the MeSH homepage, and the browser offers an online search of MeSH subject headings. This site is handy for physicians doing their own searching. The site also connects the physician to the National Library of Medicine (US), and thus to an online version of MEDLINE and PubMed central.
"The Cochrane Library consists of a regularly updated collection of evidence-based medicine databases, including The Cochrane Database of Systematic Reviews, which provide high quality information to people providing and receiving care and those responsible for research, teaching, funding and administration at all levels." This site is available on a subscription basis.
Certain features of this site require registration. Initially, it seems as if registration is free, but digging through the FAQ section reveals that a fee is involved. The site is maintained by a private company and claims to contain thousands of professionally written articles pertaining to many disorders. MEDLINE can be accessed via this site, and it has resources for PDAs.
This site provides online health information, claiming 15 million visitors per month. The company also provides "electronic data interchange services and practice management software and services to the healthcare industry."
Blumenthal, David. "Doctors in a Wired World: Can Professionalism Survive Connectivity?" The Millbank Quarterly 80.3 (2002): 525-546.
Covell, David G., Gwen C. Uman, and Phil R. Manning. "Information Needs in Office Practice: Are They Being Met?" Annals of Internal Medicine 103 (1985): 596-599.
DeCandido, GraceAnne A. "E-Reference: Closing in on 24.7." Public Library Association Tech Notes March 2001. Online. 13 Jan. 2003.
Edmonton Public Library. Licensed databases. Online. 4 Apr. 2003.
Ely, John W. et al. "Analysis of Questions Asked by Family Physicians Regarding Patient Care."
Western Journal of Medicine 172 (2000): 315-319. Online. 2 Apr. 2003.
Gorman, Paul N. "Excellent Information is Needed for Excellent Care, But So is Good
Communication." Western Journal of Medicine 172 (2000): 319-320. Online. 2 Apr. 2003.
---. "Information Needs of Physicians." Journal of the American Society for Information
Science 46.10 (1995): 729-736.
---. "Information Needs in Primary Care: A Survey of Rural and Nonrural Primary Care
Physicians." Medinfo 10. pt.1 (2001): 338-342. Abstract. Medline. U of A Lib.
2 Apr. 2003.
Graber, Mark A., George R. Bergus, and Christopher York. "Using the World Wide Web to
Answer Clinical Questions: How Efficient are Different Methods of Information
Retrieval?" The Journal of Family Practice 49.7 (1999): 520-524.
Hersh, William R. "Medical Informatics: Improving Health Care Through Information."
Journal of the American Medical Association 288.16 (2002): 1955-1958.
Hersh, William R. and David H. Hickman. "How Well Do Physicians Use Electronic
Information Retrieval Systems?" Journal of the American Medical Association 280.15
Janes, Joseph and Joanne Silverstein. "Question Negotiation and the Technological
Environment." D-Lib Magazine 9.2 (2003). Online. 8 Apr. 2003.
Kasowitz, Abby S., Blythe Allison Bennett, and David R. Lankes. "Quality Standards for Digital
Reference Consortia." Reference and User Services Quarterly 39.4 (2000): 355-363.
Library Literature and Information Science Full Text. WebSPIRS. U of A Lib. 7 Apr.
Mittman, Robert and Mary Cain. The Future of the Internet in Health Care. January 1999. Online.
7 Apr. 2003.
Sprague, Mary W. and Linda DeMuro. "Urban Physicians: Information Access." Medical
Reference Services Quarterly 15.3 (1996): 23-40.
Straw, Joseph E. "A Virtual Understanding: The Reference Interview and Question Negotiation
in the Digital Age." Reference and User Services Quarterly 39.4 (2000): 376-379.
Library Literature and Information Science Full Text. WebSPIRS. U of A Lib. 7 Apr.
Westberg, Edward E. and Randolph A. Miller. "The Basis for Using the Internet to Support the
Information Needs of Primary Care." Journal of the American Medical Informatics
Association 6.1 (1999): 2-25. Online. U of A Lib. 2 Apr. 2003.
Additional Works Consulted
Coiera, Enrico. Guide to Medical Informatics, the Internet and Telemedicine. London: Chapman
& Hall Medical, 1997.
Lankes, David R., Charles R. McClure, Melissa Gross, and Jeffrey Pomerantz. Implementing
Digital Reference Services: Setting Standards and Making it Real. London: Facet, 2003.
Smith, Roger P. and Margaret J.A. Edwards. The Internet for Physicians, 2nd ed. New York: Springer, 1999.