Journal reading patterns and preferences of pediatricians

Abstract

Purpose: To describe the journal reading patterns of pediatrician members of the American Academy of Pediatrics (AAP) and compare results to similar surveys of medical faculty and physicians. The research also explored factors that might influence changes in reading patterns in the future, such as adoption of PDA technology.

Methodology: A random sample of 2,000 AAP members was drawn from the AAP membership list, with paper surveys distributed in mid-2004.

Settings/Subjects: Six hundred sixty-six pediatrician AAP members participated in a survey of reading behavior, with a total of 1,351 members answering some questions about technology use.

Results: The hypotheses that pediatricians read many journal articles each month, read each article on average quite quickly, read heavily from personal subscriptions, read from both print and electronic journals, and read for many purposes were all supported. Pediatricians read journal articles primarily for current awareness and most often rely on quick reading from print journals for current awareness. Reading for research, writing, and presentations are more likely from library-provided electronic journals.

Discussion/Conclusion: Convenience and purpose of reading are key factors that explain reading patterns of pediatricians. Print personal subscriptions are convenient for current awareness reading, while electronic journals systems are convenient for reading for research because they provide access to a broader range of journals. Publishers and librarians must understand the purposes and patterns of reading to design appropriate journals and services. Pediatricians read many current articles very quickly and from many different locations. Pediatricians under the age of thirty-five are more likely to use PDAs, suggesting that articles delivered to a handheld device might be accepted as convenient in the future.

Highlights

• Pediatricians in practice read on average 145–184 articles per year. This is less than medical faculty rates reported in other studies but demonstrates the importance of journal articles to medical practitioners.

• Most article reading by pediatricians is for current awareness, but they also read heavily for treatment/ diagnosis and other purposes.

• Pediatricians read quickly—on average only twenty-two minutes per article.

• The majority of articles read by pediatricians are from personal print subscriptions.

Implications

• Pediatricians under age thirty-five are more likely to use PDAs, and, although few use them yet for journal article reading, they will likely be convenient for reading in the near future.

• Readings for research and readings rated as more valuable to purpose are likely to come from library electronic collections.

• Publishers must continue to design journals that are convenient, portable and allow quick reading.

INTRODUCTION

Journal reading studies over the years consistently show that physicians' journal reading behavior is different from that of other subject experts. Physicians tend to read more articles on average but spend less time per reading than do professionals in other disciplines such as engineering, astronomy, and other sciences and social sciences. Physicians also read from far more print journals (in particular, personal subscriptions) and read more frequently for current awareness than do other professionals [1]. Although physicians read many articles from personal print subscriptions, they nonetheless do use electronic journals, especially those provided by medical or university libraries.

Like other physicians, pediatricians face increasing pressures to read more, keep up with more topics and more literature, and make the best use of their time. Peer-reviewed journals and journal articles that are delivered and designed in ways that best serve these busy readers help physicians make more productive use of their limited time. In 2004 the American Academy of Pediatrics (AAP), with Carol Tenopir and Donald W. King, conducted a survey of pediatrician members of AAP. The AAP is a professional organization of approximately 60,000 pediatricians, primarily in North America. Almost all members are working physicians from a variety of practice settings including universities, hospitals, clinics, and private practice. The results of this survey will help medical publishers design better journal delivery methods and provide medical librarians with insights into the reading patterns and preferences of their constituents in order to better serve their needs.

The researchers hypothesized that reading patterns of pediatricians would closely mirror those of other physicians in earlier studies and would continue to differ from those of scientists. Changes due to adoption of information technology and wider availability of electronic journals should have an impact on reading patterns and preferences, however. This paper presents results from the 2004 survey that address reading patterns in print and electronic sources and the purpose and value of reading by pediatricians.

BACKGROUND

Studies from the 1960s to the present consistently show that medical faculty and other health professionals rely on peer-reviewed journals as a major source of quality information and read more from personal print subscriptions than from library-provided print and electronic journals. Several studies have found that physicians rely on personal subscriptions, print resources, and interpersonal communications for much of the information they need [1–7]. All of these studies suggest that physicians make use of readily available, relevant resources that require minimal time and money to obtain and that are convenient. Physicians do use libraries sometimes to locate medical articles but go to their personal collections or colleagues' collections as the first choice to keep current [2, 5–7].

Still, healthcare professionals consistently report that peer-reviewed journals are their preferred source for reliable information, followed by consultations with professional colleagues, association meetings, continuing education courses, and pharmaceutical representatives [5]. Almost all health professionals use medical journals, from an average of approximately 5 hours per month (i.e., 60 hours per year) in the 1980s [5] to an average of 118 hours per year spent reading an average of 322 journal articles, 20 years later [1].

When relevant journal articles are readily accessible, use may be even higher. Some studies have identified difficulties that physicians have using journal literature, which may pose impediments to use. These issues include: the fact that papers relevant to specific clinical issues are widely scattered across journals with different subject boundaries; the time required to search for and retrieve articles; and the time required to sift through the retrieved literature, much of which is not relevant to clinical problems [8].

In a survey of primary care physicians and physician opinion leaders (mostly academic faculty), two-thirds of the office-based practitioners and half of the opinion leaders said they found the volume of medical literature to be unmanageable. The major problems identified were a lack of time to search for information and a large amount of irrelevant material that must be screened in order to locate the desired information. Eighty-seven percent of office-based practitioners said that most physicians have at least moderate difficulty with inadequacies of terms used in Medline (Index Medicus) or other journal indexing and abstracting services [9]. These problems are further complicated by the fact that the primary method used by medical faculty to locate articles is by browsing their personal print subscriptions [1].

Medical faculty read from print proportionally more then other science faculty. Medical faculty prefer print for personal subscriptions, but more of the articles they read from the library come from electronic journals, and when a medical librarian shows them how to use electronic journals, that use goes up [10, 11]. As computer proficiency increases among medical care staff, the use of online journals will probably continue to increase. De Groote and Dorsch found a major decrease in the use of print journals when online core medical journals were introduced [11].

Sathe, Grady, and Giuse found that that clinical/research faculty use print journals more than any other type of user (i.e., fellows, nurses, residents, or nursing students) [12].

In this same study, users reported that print journals were used for reading articles and scanning contents, and that print formats contained higher-quality text and figures. Electronic journals were judged to be more easily searchable than print journals, but many articles were printed out for final reading. Medical faculty's preference for print is likely linked not only to the high volume of personal subscriptions they receive, but also to the convenience of reading print on the run [1].

D'Alessandro and Kreiter also examined information-seeking behavior of general pediatricians. Despite the fact that pediatricians' most frequently used method for information seeking is paper resources, there is a significant increase in computer usage when information seeking [10]. Physicians in general are not computer phobic. There is an opportunity for change in information-seeking behavior now that an estimated 40–50% of all U.S. physicians use personal digital assistants (PDA), although very rarely for journal or database access as yet [13, 14].

In launching the current study, the investigators hypothesized that practicing pediatricians who are members of AAP would read from personal subscriptions even more than medical faculty, would obtain articles from both print and electronic journals, and would value each form for different purposes and different reasons. Investigators also speculated that younger pediatricians, especially those who use PDAs, may be more likely to favor electronic journals.

METHODS

Surveys

A random sample of 2,000 AAP members was selected by AAP from the membership database of 60,000 pediatricians. A paper-based questionnaire was sent to each of the members in the sample on May 17, 2004. The mailing included a cover letter on AAP stationery from the AAP Executive Director asking for participation, the questionnaire, a stamped return envelope addressed to the University of Tennessee, and a $1 bill as a token incentive. A follow-up letter was sent on July 12, 2004, providing a URL for a PDF version of the instrument. All correspondence came from AAP and introduced the lead researchers (CT, DK), who received the return questionnaires and input and analyzed the data. Responses received through August 2004 are included in the analysis.

The survey, Article Reading Survey for Pediatricians (“Article Survey,” see Appendix; online only), used a form of critical incident technique, focusing on the last article read by each respondent. Many of the same questions were asked in a series of surveys of other subject specialists conducted over 3 decades by Tenopir and King [15, 19]. By asking the same questions in this questionnaire as were asked in the other survey instruments, the researchers were able to make both latitudinal and longitudinal comparisons. A second questionnaire sent to a separate random sample of 2,000 AAP members asked questions specific to Pediatrics, AAP's core journal. That survey is not discussed here, but the instrument is available at <http://web.utk.edu/~tenopir/research/aap/AAP_v19.pdf>.

Three main types of questions were asked in the article reading survey: 1) General Recollection, 2) Critical Incident, and 3) Demographic. All of the questions have stood the test of time in many previous studies [15, 19]. General recollection questions, such as questions about the number of scholarly articles read in the previous month, were kept to a minimum and to a short time frame to improve validity of responses. The questions focused on asking respondents to reflect on the last article read to give details on source, form, value and purpose of the reading (as detailed in Questions 2 through 13).

Demographic questions allow results to be analyzed by such factors as age, sex and workplace of respondents, in addition to factors of special interest to the medical community such as PDA use. Demographic characteristics of the respondents were compared to the overall demographics of the total AAP membership of 60,000.

RESULTS AND DISCUSSION

The survey provides insights into the reading patterns and preferences of AAP-member pediatricians. Taken together, amounts of reading, time spent reading, preferences for print and electronic features, purposes of reading, and sources of readings suggest ways publishers and librarians can best meet the scholarly reading needs of pediatricians.

Limitations

As with any paper-based survey, the respondents may have been more likely to be those who read and respond to paper mail. Perhaps respondents might therefore be more conservative than the population as a whole. Respondents were also more likely to be from the U.S., as researchers offered a stamped return envelope to those members.

Many of the concepts and questions asked in the questionnaire are complex. Concepts such as reading were defined, but others such as interlibrary loan or online databases were not. The investigators assumed that given a comprehensive list of choices, respondents would find the one best answer or add an “other” choice. This complexity may have reduced the response rate.

Demographics

The questionnaire was returned by 666 respondents (33.3% return rate). The respondents were characteristic of AAP membership in the demographic factors of age and gender, but were more often in private practice or hospitals/clinics and less often in group practice than the overall AAP membership (see Table 1). These practice settings are not likely to result in major differences in reading patterns, but the higher percentage of respondents in an academic setting likely means that the respondents read more for research on average than do AAP members as a group, as clinicians typically do less research than academic faculty.

Table 1 Demographic comparisons between survey and AAP membership

Amount of reading

Pediatricians estimated they read on average between 145 and 184 articles per year (the lower figure eliminates 2 outliers who reported reading 999 and 1,000 articles each month). This finding is similar to results found in earlier national surveys of physicians [19]. In all of the investigators' surveys, reading is defined as going beyond the table of contents, title, and abstract to the body of the article.

This average number of readings is much lower than that of medical faculty at the University of Tennessee, who report reading on average approximately 322 articles yearly, but consistent with other surveys exclusive to medical faculty [1]. Two factors—1) workplace and 2) work responsibilities—likely account for relatively low amounts of overall reading by practicing pediatricians, only 20% (134 of 649 responses) of whom were university faculty and almost all of whom have significant clinical responsibilities, as compared to full-time medical faculty. Many studies have observed that professionals outside the university read considerably less than do university faculty [15]. In 2001– 2003 surveys at 3 universities, scientists (representing all fields) on average reported reading 216 articles yearly [16].

In addition, respondents with research and writing responsibilities tended to read more than those with heavy administrative or clinical responsibilities. AAP members are primarily medical practitioners. Respondents were asked to indicate the proportion of their time spent on various responsibilities (e.g., teaching, clinical practice). Over a quarter of the total respondents (168 of 655) reported spending 100% of their time on medical practice. Three quarters of the respondents (n = 527) spent more than half of their time on medical practice, and more than one-third (n = 283) spent 95% or more on medical practice. Respondents were not researchers—about three-quarters (n = 438) spent under 5% of their time on research. Teaching and administration took up even less of their time.

Age of articles read

In studies of scientists and social scientists from 1977 to 2003, researchers have found consistently that, on average, across all fields, approximately two-thirds of readings are from the current year of publication, 20– 23% are 2 to 5 years old, and the remaining 12–14% of readings are older than 5 years [17]. For pediatricians, the dates of articles last read range from 30 years old to the current year (1974 to 2004), but 81% (517 of 636 responses) of the readings were from the current year (2004 at the time of the survey) (Table 2). Other fields of science or social science and university faculty across fields report a higher percentage of older readings [17]. This is not surprising because medical readers, in particular those with clinical responsibilities, tend to read heavily for current awareness purposes. Not surprisingly considering the preponderance of current year readings, 87% (575 of 661) of the respondents said they had never read the particular article before, yet over half (351 of 637) of the pediatricians indicated that they knew “about the information reported or discussed in this article” prior to their first reading of the article.

Table 2 Year of last article read

The influence of personal demographics on reading

Gender and age made some difference in the number of scholarly articles that pediatricians read. The male participants in the survey read a statistically significantly higher number of articles each month than did the female participants (an average of 15.6 articles/ month for males compared to an average of 9.5 readings for females; t = 4.444, P < .0001). Respondents who were 35 or older (443 respondents) read 13.8 scholarly articles in the past 30 days on average, significantly more than their younger counterparts (174 respondents), who read only 8.2 articles on average (t = 5.251, P < .0001); however, when the age cutoff was 40, there was no significant difference. Since female pediatricians are, on average, younger than male pediatricians, this factor may explain the gender differences as well.

The increased prevalence of reading after age 35 may be because younger respondents have just recently completed their medical training and feel more up-to-date with the current literature. The amount of reading of scholarly articles did not seem to vary with additional advanced degrees (such as a PhD degree) earned by pediatricians.

Studies of other groups show a clear relationship between amount of reading and awards or other special recognition [15]. This relationship also holds true for pediatricians in this study. Respondents to the survey who received awards or special recognition for research or other profession-related contributions in the past 2 years read statistically significantly more scholarly articles (mean = 18 readings per month) than those who had not (mean = 10.1, t = 5.352, P < .0001).

Time spent

Pediatricians, like other physicians, read quickly. On average, pediatricians reported they spent 20 minutes per reading (Table 3). At an average of 20 minutes per reading and an estimated 145 or 184 readings per year, pediatricians spend an average of 49–61 hours per year reading articles. This does not take into account the time spent searching for relevant articles to read nor the time to apply what they have learned in the reading.

Table 3 Average time spent reading per article

Since they read fewer articles than do researchers and read each quickly on average, clearly the total time spent reading per year is lower for pediatricians than for scientists (aggregated, all disciplines), full-time university medical faculty, astronomers, and many other groups. Still, a substantial investment of busy practitioners' time is spent reading journal articles, primarily for keeping current in their profession.

Outcomes of reading (novelty and value)

Subject experts read journal articles for many reasons. Pediatricians read most often to keep up with their field—the principal purpose of reading for 50% (313 of 626 responses) of readings by pediatricians was current awareness. Pediatricians are practical readers who need current and highly relevant materials to do their job. Time spent reading and source of articles both reflect this skewing toward reading for current awareness.

Respondents were allowed to select as many secondary purposes as they chose. The most frequently selected secondary purposes were: treatment (30.3%, 202 of all 666 respondents), current awareness (29.3%, 195 of 666), and diagnosis (20%, 133 of 666). Teaching was a close fourth (18.2%, 121 of 666), but no other purpose was selected by more than 10% of all respondents. Reading for research, proposal writing, or presentations were quite rare among AAP respondents.

Respondents selected many different ways in which their reading affected the principal purpose for reading an article. Reading most often helped improve a result (43.5%, 290 of all 666 respondents), inspired new thinking (42.2%, 281 of 666), and helped narrow the focus (17%, 113 of 666). All other outcomes were selected by less than 5% of respondents. Scholarly article readings were important to achieving each reader's principal purpose: “somewhat important” to 71.5% (462 of 646 respondents to this question) and “absolutely essential” to 26.6% (n = 172) (Table 4).

Table 4 Rank order of outcome of reading*

Associations between the principal purposes of reading and source types were found using Chi-square tests. The source types were regrouped in two ways: 1) personal, library, and others (χ² = 125.782, P < .0001), and 2) print, electronic, and others (χ² = 75.483, P < .0001). Compared to the other principal purposes (Table 5a), pediatricians reading for research, writing, and/or presentations tended to access articles less from personal sources (31%, 18 of 58), significantly lower than the percentages for teaching, reading for current awareness, and other purposes, but more from libraries (50%, 29 of 58). Pediatricians reading scholarly articles for current awareness significantly differed in that they seldom obtained articles from a library source (3.3%, 10 of 301).

Table 5a Association between principal purpose and source

Pediatricians reading scholarly articles for research, writing, and/or presentations tended to rely more on electronic sources (55.2%, 32 of 58), significantly more than they used electronic sources for other purposes, such as current awareness, teaching, and consulting (Table 5b).

Source of reading

Unlike many other fields of science or social science, physicians still rely on personal print subscriptions for many readings (Table 6; online only) [1, 15]. Over 68% of the last article readings by AAP members came from personal print subscriptions (433 of 632). Readings from personal electronic subscriptions were relatively rare, however: only 3.2% of total readings (n = 20). Readings from library subscriptions were most likely to be from electronic journals, with only 1.7% (n = 11) of readings from print subscriptions, but 6.3% (n = 40) from library electronic subscriptions, consistent with findings from previous other surveys by Tenopir and King. Questions 6, 7 and 9 in the questionnaire provided a detailed list of choices for source and form of reading (see Appendix; online only).

One reason for this behavior may be library collection development policies that favor electronic journals, but scientists and social scientists in many studies also say they prefer electronic journals when using library-provided journals. Overall, however, approximately 81.3% of readings by pediatricians in this study came from print resources of one kind or another (514 of 632 reported readings). Approximately 16% (101 of 632) came from electronic sources, with 2.7% (17 of 632) unknown (other). This is almost exactly the reverse of astronomers, who report nearly 80% of readings from electronic sources, including electronic journals, websites, and e-print servers [18].

Even when the delivery mechanism is electronic, paper is still the reading format of choice (Table 7; online only). Only 47 respondents reported reading from a computer screen (7.3% of 644 respondents who answered this question). All of the other respondents read the article on paper. Previous studies have reported a number of reasons why readers prefer paper, including convenience, quality, ease on eyes and body, portability, and ability to write on easily [16].

Since a high percentage of readings came from personal print subscriptions and were of current articles, it is not surprising that 70.1% or 411 of the 586 respondents to this question found the article by browsing in a personal print subscription. In total, browsing accounted for locating 76.5% (n = 448) of the articles. Searching was used to find 8.2% (n = 48) of the articles, most commonly in a scientific indexing/abstracting database (5.5%, n = 32). Seven percent (n = 45) were found when a colleague “told me about it.” Only 2.8% (n = 16) of readings were located through citations, and 4.9% (n = 29) from “other” (e.g., attending a conference or a journal club) sources.

Factors that may influence the future

Availability of information technology, both within the home and via mobile technologies, may make reading electronic journals more convenient in the future. Availability of technology may influence location of reading, the necessity of print personal subscriptions, and other reading behaviors. Gathering information on these factors now may serve as a predictor, or at least a baseline, for future reading patterns.

Location of reading

Pediatricians do their scholarly reading most often at home—50.5% or 328 of the 650 respondents who answered this question indicated they were home when they read the critical incident article. The second most popular location was at the office (36.8%, 239 of 650), with hospital/clinic far behind (6.2%, 40 of 650). Other places, including a library, in transit, at the gym, on the beach, or on an airplane, accounted for 6.7% (43 of 650) of the readings. The investigators did not ask questions about technology at home, such as high-speed Internet connections, but it is a relevant question for future surveys due to the high proportion of reading from home.

Subscription payment

Respondents reported receiving a varying number of personal subscriptions that they paid for themselves (mean = 2). In addition, some respondents (221) received subscriptions for personal or departmental use paid by a grant. The final category of subscriptions was free to the reader, with approximately two-thirds (422 of the 633 respondents) reporting they receive free subscriptions.

Taking all subscriptions together, pediatricians average slightly over 5.3 subscriptions each. The researchers' most recent survey of full-time medical faculty found they averaged just under 6 subscriptions each [1]. In both cases, these are far above the average number of subscriptions for scientists and social scientists. In the most recent surveys, the average number of subscriptions for scientists and social scientists in universities, government laboratories, and corporations was down to just under 2 per person, a steady decline since the 1970s [19].

PubMed use

Most respondents had used PubMed at some time (70.9%, 461 of 650 respondents who answered the question). Of those 461 who answered yes to ever having used PubMed, just 422 reported the number of times in the past 12 months. On average, respondents who used PubMed estimated they used PubMed 33.6 times in the past 12 months. As a result of using PubMed, respondents estimated they read from between 0 and 1,000 articles in the last month (mode = 10, mean = 14.6, with outliers of 999, 1,000, and 1,500 omitted).

PDA use

Because 46.7% of respondents (307 of 657) noted that they did not use a PDA, the investigators concluded that slightly over half did use a PDA. This finding aligns with recent surveys that show that about 50% of U.S. practitioners use handheld devices and 80% of U.S. medical students use them [13].

Very few AAP members use PDAs as yet to access patient information (6.1%, 40 of 657) or to download articles (2%, 13 of 657). This may reflect current availability and habits rather than predict future behaviors, however. As more scholarly literature becomes available in a PDA-friendly format, some pediatricians may take advantage of this convenient format.

Age is significantly associated with whether a person uses a PDA or not. In the Chi-square test (χ² = 15.125, P = .001), investigators used the age 35 as a cutting point. Participants who were older than 35 were less likely to use a PDA: 50.1% of such respondents (229 of 457) indicated, “I don't use a PDA.” In contrast, only 33.5% of participants (64 of 191) who were 35 or younger did not use a PDA.

As far as the reasons for using a PDA are concerned (Table 8; online only), significant differences were found by age for the reasons “to access reference material” (χ² = 13.769, P = .001) and “for information services” (χ² = 18.949, P < .0001). Among those using a PDA, younger people were significantly more likely to use a PDA to access reference material (75.6%, 96 of 127) and for information services (72.4%, 92 of 127) than people older than 35 (56.1%, 128 of 228; 48.7%, 111 of 228).

The team also found that practice setting is significantly associated with the use of a PDA (χ² = 16.977, P = .005). Compared to the other settings, respondents in hospitals (60.6%, 42 of 691) and universities (63.2%, 133 of 210) were more likely to use a PDA, whereas respondents in HMOs were less likely to (33.3%, 12 of 36). No evidence showed that practice setting was associated with why people use a PDA.

Pediatricians remain rather traditional in their journal reading habits. PDAs were not (as yet) being used for article reading. In addition, only 6% (41 of 683) of respondents indicated they had ever downloaded an article from a pay-per-article source. Of those 41, only 25 gave a price. The median and mode price (7 of 41 respondents) was $10.00, with a mean price of $12.83. Payment prices ranged from $2.50 (2 responses) to $30.00 (3 responses.)

Conclusions and implications for the health community

The hypotheses that pediatricians would read heavily from personal subscriptions, would read from both print and electronic journals, and would read for many purposes were all supported by this study of AAP members. Respondents' reading from personal print subscriptions was even more prolific than anticipated, however. Pediatricians, like other physicians, read journal articles primarily for current awareness. They used formats and features that make it convenient to quickly keep up with new developments, thus relying on print journals and personal subscriptions for the bulk of their readings. Electronic journals, while useful for some purposes, were still not as convenient for this group as a printed journal landing in the mailbox. It may therefore be prudent for publishers to continue to invest in updating printed journals for physicians, ensuring that they can be easily read and navigated. All of these participants receive the printed version of Pediatrics as well as access to the electronic version as part of their AAP membership. Clearly, they find the print version more convenient for current awareness reading.

Although current awareness was the principal purpose for a majority of readings, it was not the only reason pediatricians read articles. When they read for research, writing, or presentations, they were more likely to read from library-provided electronic subscriptions. Librarians may therefore wish to consider that, while current awareness may be the principal driver of journal reading in print, most physician library users will be seeking specific information from electronic sources (that they ultimately will want to print out).

The hypotheses that younger pediatricians were more likely to use PDAs and favor electronic journals were partially supported. Many pediatricians embrace technology such as personal digital assistants when it helps them do their work. The fact that PDA use and use of the electronic journals is higher among pediatricians younger than thirty-five may indicate that a changing relationship with print and online journals within the field is on the horizon, but there is little reading on PDAs as of yet. This will only occur if electronic journals are seen as convenient by physician readers or if practice settings and schedules change to better facilitate online journal reading.

For all scientists, the purpose of reading determines what format is likely to be more convenient. Convenience and purpose are key factors for publishers and librarians to understand in order to provide what pediatricians need, in the formats they need it. Medical librarians and publishers need to provide products and services in the form and manner of most use to their constituents. Pediatricians have many alternatives for the scholarly journal articles they need. Although they are not computer phobic, they choose the forms that are most convenient to them, which today are often still printed personal subscriptions for current awareness, but electronic library versions for research. Print journals are likely to be important to medical practitioners for some time to come, but changes may be coming, as younger pediatricians are more likely to use both e-journals and PDAs.

Table 5a Association between principal purpose and source