The informationist: building evidence for an emerging

health profession*
Suzanne F. Grefsheim, MEd, MSLS, FMLA; Susan C. Whitmore, MS, MSLS;
Barbara A. Rapp, PhD, MLS; Jocelyn A. Rankin, PhD, MLn, AHIP, FMLA;
Rex R. Robison, PhD, MLS; Candace C. Canto, MA, MLS
See end of article for authors’ affiliations. DOI: 10.3163/1536-5050.98.2.007

Background: To encourage evidence-based practice, receiving the service. Survey data were analyzed with
an Annals of Internal Medicine editorial called for a new Pearson’s chi-square or Fisher’s exact test.
professional on clinical teams: an informationist
trained in science or medicine as well as information Results: Comparing 2006 to 2004 survey responses,
science. the researchers found that study participants
reported: seeking answers to questions more
Objectives: The study explored the effects of frequently, spending more time seeking or using
informationists on information behaviors of clinical information, believing time was less of an obstacle to
research teams, specifically, frequency of seeking finding or using information, using more information
information for clinical or research decisions, range of resources, and feeling greater satisfaction with their
resources consulted, perceptions about access to ability to find answers. Participants’ opinions on
information, confidence in adequacy of literature informationists’ qualifications evolved to include both
searches, and effects on decision making and practice. subject knowledge and information searching
It also explored perceptions about training and expertise.
experience needed for successful informationists.
Conclusion: Over time, clinical research teams with
Methods: Exploratory focus groups and key informationists demonstrated changes in their
interviews were followed by baseline and follow-up information behaviors, and they valued an
surveys conducted with researchers and clinicians informationist’s subject matter expertise more.

INTRODUCTION Highlights
In their 2000 Annals of Internal Medicine editorial,
Davidoff and Florance called for a new role on the
N Informationist involvement in traditional team activi-
ties—going on rounds and searching and critically
clinical care team: the informationist [1]. This role was evaluating the literature—increased over time.
needed, they believed, to bring evidence to clinical
practices facing continued growth of published N As the relationship between a clinical team and
informationist developed, activities expanded to
literature, patient safety concerns, and general lack
include projects such as development of wikis,
of time available to health care professionals. The
Institute of Medicine (IOM) also noted that training databases, and websites.
and encouraging clinicians to identify and apply N Clinical teams came to view subject knowledge as
evidence was not the complete solution to improving key to an informationist’s preparation; however, their
practice [2]. They too suggested that an informationist expertise as information scientists was valued most
be part of the clinical team. Davidoff and Florance and highly.
the IOM thought clinical knowledge and experience, N ‘‘Initiative,’’ approaching research staff in their work-
as well as strong information science and related place, was the one personal trait focus group
technology skills, were required to perform this participants agreed on as most desirable for an
function. informationist.
The National Institutes of Health (NIH) Library
hired its first two informationists—librarians with
N When first introduced to the concept, researchers
cited customization of services to team needs as a
extensive expertise in a clinical or research specialty— major benefit.
in 2001 and assigned them to research teams in their
fields of expertise. Over the intervening years, the
Implications

* This research was supported in part by the National Institutes of

N To assure a successful informationist program,
libraries must be prepared to commit time and
Health Director’s set-aside funds for program evaluation and by the
National Library of Medicine.
money.
N Whether informationists or not, librarians should be
This article has been approved for the Medical Library prepared to support users’ increased need for
Association’s Independent Reading Program ,http://www.mlanet genetics and molecular biology information.
.org/education/irp/..

J Med Libr Assoc 98(2) April 2010 147

Grefsheim et al.

informationist program grew, and currently the informationist maturation models: First is an embedded
library has fifteen informationists who are members model, usually in a clinical setting, that initially focuses
of more than forty NIH clinical and/or basic science on traditional information services but later progresses
research teams. In addition to being expert informa- to support the team’s technical and informatics infor-
tion scientists, some of the NIH informationists have mation needs as well. Second is the consultant or
doctoral degrees in basic biomedical sciences such as bioinformatics model, which begins with a strong
molecular biology, biochemistry, and neuroscience. technical focus and, over time, provides more person-
All of them are expected to continue their education alized service [17]. While a few institutions were true
by taking graduate-level courses in the specialties innovators, the literature review also confirmed that
they support and participating in the specialties’ informationist programs, when considered as innova-
professional meetings, in addition to maintaining tions, remained in the early adopter stage, with libraries
their skills as information scientists. (A more detailed trying out the idea ‘‘in a careful way’’ [18].
description of the NIH informationist program was Findings from studies in clinical settings suggest
published in 2008 [3].) From the start, informal that embedding informationists encourages question-
feedback from the research teams was positive; ing and that trained informationist librarians can
however, as the commitment to the program in- perform critical appraisal of the literature comparable
creased, a formal evaluation was suggested. The to clinicians [14, 19, 20]. These studies also show that
study reported here looks at the effects the NIH program success requires technical expertise, service
informationists have had on participating clinical excellence, and commitment by management [3, 13].
research teams. In addition, success requires informationists to dem-
Because every patient at NIH is enrolled in a clinical onstrate domain knowledge, engage in continuous
trial that takes place over an extended period with learning, and be completely embedded in the team
several return visits and the costs of participation— [3]. Studies of the bioinformatics informationist
including transportation, lodging, and meals—are model, on the other hand, have been descriptive but
covered by NIH, traditional measures of information suggest that subject knowledge (molecular biology in
effects in a hospital—such as differences in patient particular) in support of project-specific consultation,
length of stay, money saved, or patient outcome— generalized training, and development of web portals
would not provide the data needed to evaluate the NIH for easy access to sequence analysis tools and other
Library’s informationist program. As a result, the resources are key elements for success [11, 12].
evaluation looked instead at changes over time in the Evaluation of programs employing both informa-
information behaviors of the participating clinical tionist models is needed but is challenged by the fact
research teams, not specific individuals. The training that these programs are inherently customized and
and experience needed to perform the informationist targeted to small groups [17]. However, it is still the
role effectively were also explored. case, as Schacher observed, that while the benefits of
Numerous studies exploring the information be- having the literature available at the point of care are
haviors of health professionals have underscored the clear, more and better data on the impact of
need for better support for their information needs. informationists are needed to secure routine position-
Covell found that many information needs were not ing of these professionals on health care teams [21].
recognized by practicing physicians and others went This study attempts to provide evidence of the impact
unanswered [4]. In 1991, analyzing information that NIH’s informationists have had on the clinical
requests during clinical teaching, Osheroff reported research teams with whom they work.
frequent requests, but many required the synthesis of
patient information and medical knowledge and thus STUDY QUESTIONS
were difficult to answer [5]. A more recent taxonomy
of fifty-nine barriers to information use illustrated the Specifically, this study attempted to answer the
complexity of developing successful interventions [6]. following questions:
However, Gorman showed how the biomedical 1. Does the presence of the informationist:
literature of 1994 could answer primary care infor- a. increase the frequency with which teams seek
mation needs, and, more recently, Westbrook found information to support clinical or research decisions?
that use of online information resources enhanced the b. increase the range of information services consult-
accuracy of answers provided to typical clinical ed in response to information needs?
problems by experienced clinicians [7, 8]. Nonethe- c. facilitate and improve access to information rele-
less, online databases and libraries, even when vant to clinical practice?
available, continue to be infrequently consulted due d. increase the confidence of clinical research teams
to lack of time and search skills [9]. that they have adequately researched the available
The 2000 Annals of Internal Medicine editorial proved published literature?
a stimulus to librarians already exploring ways to e. improve clinical and research decision making and
better integrate information into the clinical context. practice through enhanced access to the published
In the United States and internationally, academic literature?
medical centers and other health organizations estab- 2. From the clinician’s perspective, what education,
lished model informationist programs [3, 10–16]. A experience, and personal characteristics are important
recent literature review found what appears to be two contributors to a successful informationist?

148 J Med Libr Assoc 98(2) April 2010

 The informationist

METHODS To encourage survey response, lead researchers and

other principal contacts from the participating clinical
Study site research teams were asked to send their teams a link
to the web survey. While the survey was underway,
NIH not only funds translational clinical and basic
when an informationist joined a new team, its
research through extramural grants and contracts, but
members were asked to complete the survey; ulti-
also conducts laboratory research and hundreds of
phase I or II clinical trials in its own intramural mately, individuals from 9 teams could respond to the
research program. Eighteen of the 27 NIH institutes survey (,150–200 individuals). A total of 74 surveys
and centers have intramural research programs. Most were completed by February 2005, when the baseline
NIH clinical trials take place at the Bethesda, survey ended. Respondents were anonymous and
Maryland, NIH Clinical Research Center, a 242-bed, were prevented from answering more than once. A
in-patient hospital with 90 day-stations for out-patient definitive response rate is unknown, because the
visits. Ninety-five percent of the researchers working actual number receiving the survey was determined
in these laboratories and clinics reported seeking by the research team leaders and not shared with the
information themselves, and 91% reported they study coordinators. However, given the size of the
preferred doing so according to a 2005 information potential survey population (permanent staff and
needs assessment [22]. Institutional review board fellows who rotate in and out every few months),
approval was not required for this study, because the estimated response rate was 40%–50%.
NIH does not require it for program evaluations Initially, an attempt was made to use the new teams
conducted among its staff by central service organi- as a control group, so that comparisons could be made
zations such as the NIH Library. between groups who had worked with information-
ists for a year or more and those who had never
worked with one. The intention was that new teams
Focus groups
would complete the survey before the informationist
The study began with exploratory work to identify came on board. However, team members were slow
and describe NIH researchers’ expectations, percep- in responding. Ultimately, the time allowed for
tions, and experience with the informationist concept. completing the survey meant that some or all
In October 2003, a qualitative research consultant members of every team had worked with an
conducted three ninety-minute focus groups with ten informationist before they responded to the survey.
scientists, representing both clinical and laboratory Consequently, no control was established, and valid
staff, from one institute. About half had worked inter-group comparisons were not possible.
closely with an informationist. Topics included
discovering new ways that an informationist might Follow-up survey, 2006
contribute to their work, gathering information to
enhance the informationist role, and informing the The same survey was repeated during 4 weeks from
design of the planned quantitative study that would mid-May to mid-June 2006. Because of the time
reach a larger group receiving informationist services. required to obtain responses to the baseline web
Using the same semi-structured interview guide, the survey (13 months), the follow-up survey was
consultant also conducted three key informant inter- conducted by telephone by a consultant who special-
views with the same institute’s leadership. Focus ized in telephone surveys. To give the consultant an
groups and interview data were transcribed, and appropriate ‘‘call list,’’ the informationists were asked
content analysis was conducted by the consultant to to identify all members of the clinical research team
identify themes and patterns. with whom they worked. Study coordinators validat-
ed the final list. One hundred seventy members of the
Baseline survey, 2004 9 original participating teams were sent an email
inviting them to participate in the survey. Eighty-four
Following the focus groups and key informant responded, a 49% response rate.
interviews, the NIH Library contracted with a market
research and consulting firm to develop and then Survey population
implement a survey of NIH staff who had, or were
about to have, an informationist on their team. The All NIH clinical research teams consist of a principal
focus group findings informed the content and format investigator (PI), one or more co-PIs, and several
of the survey. For example, the web survey format postgraduate research fellows, who spend several
was chosen because it offered flexibility for the weeks or months with a team as part of their rotation
clinical researchers being studied, something noted through various intramural research groups in an
as particularly critical by focus group participants. institute. About half of all fellows at NIH are foreign
The survey method allowed for two iterations: a nationals and are at NIH for two to five years as part
baseline and a follow-up survey eighteen to twenty- of an international training program. PIs and co-PIs
four months later. In January 2004, the survey was have medical degrees (MDs) and/or doctoral degrees
pretested by a small group of clinical researchers who (PhDs) and are the largest segment of a clinical
worked with an informationist. The final survey was research team. Fellows assigned to clinical research
posted to the web in February. teams typically have PhDs; a few have MDs. They

J Med Libr Assoc 98(2) April 2010 149

Grefsheim et al.

Table 1
Comparison of the two surveys
Baseline survey 2004 Follow-up survey 2006

Format Web Telephone

Duration February 2004–February 2005 May 2006–June 2006
# asked to participate Estimated 150–200 170
Response rate Estimated 40%–50% 49%
Informationist tenure 66 months–3 years 3–5 years
Content Developed by consultant based on focus groups and key informant interviews Same questions as 2004
Demographics (n574) (n584)
& 21% principal investigators (PIs) and co-PIs (16) & 54% PIs and co-PIs (45)
& 19% fellows (14) & 13% fellows (11)
& 15% research nurses/study coordinators (11) & 4% research nurses/study coordinators (3)

make up the second largest segment of a team, but participants who had used this or a similar service
they are transient. Other members of a clinical remarked on the competence and trustworthiness of
research team can include a research nurse or study informationists and generally gave them high praise.
coordinator, a staff nurse, a statistician, a data The consultant compared this to Federal Express’s
manager, a pharmacist, and a nutritionist. early market research that indicated low interest or
Because of the nature of their appointments, the perceived need for overnight delivery. It was not until
same fellows who responded to the survey in 2004 are the service existed and people began using it that
highly unlikely to have taken the survey again in users understood its value.
2006. However, all other categories of respondents are After colleagues endorsed the concept, skeptics
permanent employees who generally remain part of were more willing to discuss the potential benefits of
the clinical research team for many years. Therefore, a librarian or informationist on their team. Many saw
the likelihood that PIs or co-PIs, research nurses, and advantages to informationists attending rounds and
the other members of the team who responded to the staff meetings. Customizing information services to
survey in 2004 also responded in 2006 is high. Because the team’s specific needs was a general preference.
the surveys were anonymous, there is no way to Participants had difficulty articulating the personal
confirm this. characteristics or training that an informationist
should have, given that most had never encountered
Survey analysis one. They were more comfortable talking about
desired skills, such as competence with technology,
Responses to the survey questions were first analyzed critical thinking, and knowledge of the scientific
for inter-year comparisons by the market research method. The one trait they did identify was ‘‘initia-
consultants who developed the survey (Table 1). tive.’’ The consultant conducting the interviews noted
Their analysis drew on their broad knowledge of the that the idea of an informationist approaching them in
information management function but did not include their workplace was ‘‘hugely appealing.’’ They
detailed statistical analysis, such as cross-tabulation of especially welcomed someone who could suggest
two or more questions, and did not include signifi- better ways to search, retrieve, and organize informa-
cance testing. To conduct the desired analyses, the tion.
authors converted the consultants’ Excel spreadsheets
into JMP statistical software files (SAS, Cary, NC). For Survey findings
example, cross-tabulations were done between the
question about the respondent’s role on the team and For both survey iterations, PIs or co-PIs, fellows, and
several of the twenty-two other questions on the nurse researchers or study coordinators constituted
survey. Responses were analyzed using Pearson’s chi- the majority of respondents. In 2004, the proportion of
square test. In cases where cell values were less than respondents in each of these 3 categories was
five, Fisher’s exact test was used (noted by {). comparable. In 2006, however, almost 3 times as
Statistics were calculated using JMP, version 7.0. many PIs and co-PIs responded (45/84) than had in
2004 (16/74), while about a fourth as many nurse
RESULTS administrators responded (3/84) than had responded
in 2004 (5/74). A similar number of fellows responded
Focus group findings, 2003 both years, 11/84 in 2006 and 14/74 in 2004. The
number of respondents in the other categories of
Despite expressing overall satisfaction with their research team members remained about the same.
information-gathering capabilities, focus group par-
ticipants reported difficulty sorting through the Information behavior
plethora of information and finding specific informa-
tion that they needed. The suggestion that an To address study questions related to whether having
informationist could be a solution to this dilemma an informationist on a clinical research team resulted
was greeted with skepticism by participants who had in information behavior changes, several questions
no experience with the program. However, the few were asked of team members about their ability to

150 J Med Libr Assoc 98(2) April 2010

 The informationist

Table 2
Comparison of information seeking from baseline to follow-up survey
2004 (n=74) 2006 (n=84)

2.1. How frequently are you able to pursue answers to questions that you think could be answered by a search of the published literature?
Less than 20% of the time 19% (14) 2% (2)
20%–40% of the time 18% (13) 7% (6)
40%–60% of the time 22% (16) 12% (10)
60%–80% of the time 26% (19) 29% (24)
80%–100% of the time 16% (12) 50% (42)
2.2. What are the influences on whether you seek the information to answer these questions?
Urgency of question 50% (37) 57% (48)
Time to look for answer 47% (35) 29% (24)
Ease of finding answer 64% (47) 70% (59)
Curiosity about answer 34% (25) 33% (28)
Other 5% (4) 10% (8)
2.3. What is your most likely source of answers?
Electronic journals or databases 69% (51) 80% (67)
Free information on Internet 11% (8) 15% (13)
All other choices (including colleagues and librarians) 20% (15) 5% (4)
2.4. What databases do you use? (multiple responses permitted) (top 10 responses in 2006 shown)
PubMed/MEDLINE 95% (70) 96% (81)
Web of Science 27% (20) 54% (45)
MD Consult 20% (15) 39% (33)
Cochrane Library Reports 19% (14) 45% (38)
GenBank/DNA sequences 16% (12) 35% (29)
EMBASE 5% (4) 24% (20)
Protein sequence databases 4% (3) 24% (20)
Other molecular biology 1% (1) 19% (16)
Biological Abstracts 8% (6) 18% (15)
Other 15% (11) 40% (34)
2.5. What are the reasons for failure in finding information?
Lack of time to search in all relevant places
Often 47% (35) 27% (23)
Sometimes 47% (35) 61% (51)
Never 5% (4) 12% (10)
Relevant information too hard to find
Often 14% (10) 10% (8)
Sometimes 76% (56) 62% (52)
Never 11% (8) 29% (24)
Insufficient training on how to search
Often 24% (18) 8% (7)
Sometimes 54% (40) 62% (52)
Never 22% (16) 30% (25)
2.6. How satisfied are you with your ability to obtain answers, compared to a year ago?
More satisfied 35% (26) 55% (46)
Similarly satisfied 41% (30) 43% (36)
Less satisfied 1% (1) 1% (1)
Not sure 23% (17) 1% (1)

find answers to questions that arose in their work that cists—as a group, their likelihood of seeking infor-
they thought could be answered by a search of the mation remained unchanged, 35% (6/17) in 2004,
literature. By 2006, 79% of all respondents reported compared to 38% (5/13) in 2006.
pursuing answers to these types of questions more Respondents also reported spending 37% more
than 60% of the time, compared to only 42% of time each week searching the literature, retrieving
respondents in the baseline survey (P,0.001) (study materials, and reading 9.2 hours per week on average,
question 1a) (question 2.1, Table 2). Much of this compared to 6.7 hours per week in 2004. Despite this,
increase was driven by PIs and co-PIs. In 2004, 56% they were significantly less likely to identify ‘‘time to
(9/16) of the PIs and co-PIs reported pursing answers look’’ as an obstacle in 2006 than they had in 2004
in the literature to questions that arose in their work (47% in 2004 to 29% in 2006) (P50.01) (question 2.2,
60% or more of the time. In 2006, 84% (38/45) of PIs Table 2). Further, respondents were as likely in 2006
and co-PIs reported doing so. Fellows also were more to seek information themselves, rather than have
likely to pursue answers that could be found by someone else do it for them, as they were in 2004: 85%
searching the literature. In 2004, 78% of fellows (11/ of the time in 2004 and 81% of the time in 2006. This is
14) reported pursuing answers to these types of a somewhat lower percentage than the general
questions 60% or more of the time; in 2006, 100% of population of NIH researchers, perhaps explained
fellows reported doing so. Looking at other team by the larger number of clinicians in the survey
members—such as research and staff nurses, data population than in the general NIH research popula-
managers, statisticians, nutritionists, and pharma- tion surveyed in 2005 [22]. Respondents were more

J Med Libr Assoc 98(2) April 2010 151

Grefsheim et al.

likely to use ‘‘electronic journals and databases’’ to ties, including ‘‘going on rounds,’’ ‘‘helping with
find information in 2006 (80%) than in 2004 (69%) searches,’’ and ‘‘summarizing and screening the
(question 2.3, Table 2). literature.’’ Further, a notable percentage (more than
When asked which databases they used to find 36% in 2006) reported their informationist was
information to answer questions that arose in their engaged in ‘‘other’’ tasks not on the list of options,
work (study question 1b), in 2004, 95% of all suggesting that informationists were participating in a
respondents reported using PubMed/MEDLINE broader array of team activities than anticipated when
(question 2.4, Table 2), and the range of reported the service began. Involvement in nearly all activities
use of other databases was 1%–27% (median 11%). By increased over time, again suggesting that informa-
2006, use of databases other than PubMed was much tionists were consulted more often and for more types
higher (range 8%–54%, median 24%). For example, of tasks. Another indicator of increased team respon-
Web of Science (P50.001), Cochrane Library Reports sibilities was that, in 2004, only 40% of all respondents
(P50.001), and GenBank (P50.01) were used signif- (30/74) reported working directly with the informa-
icantly more often in 2006 than in 2004. Large tionist as opposed to merely having an informationist
increases also were reported in the use of other assigned to their team, while, in 2006, 69% of
protein and molecular biology databases in 2006 respondents (58/84) reported working directly with
compared to 2004. the informationist.
Looking just at PIs and co-PIs, the likelihood that To measure acceptance of the informationist on the
they would consult Web of Science went from 25% (4/ team, the researchers asked those who had reported
16) in 2004 to 62% (28/45) in 2006; 19% of PIs and co- working directly with informationists if they would
PIs (3/16) reported using Cochrane to answer do it again (question 3.2, Table 3) and whether they
questions in 2004, compared to 58% (26/45) in 2006. would recommend an informationist to others (ques-
Similar increases in use of both Web of Science and tion 3.3, Table 3). Responses were highly positive on
Cochrane were noted for fellows. Use of GenBank both iterations of the survey, but even more so in 2006
increased among PIs and co-PIs, and fellows as well. than 2004, for example, with 100% in 2006 reporting
In 2004, 13% of PIs and co-PIs (2/16) and 36% of they would use the service again, compared to 80% in
fellows (5/14) reported seeking answers to questions 2004.
there. In 2006, 42% of PIs and co-PIs (19/45) and 64% Whether they reported using the service or not,
of fellows (7/11) reported using it. Similar rates of respondents were presented with a list of potential
increase were noted for other molecular biology– benefits of the program and asked to select those they
related databases. There was little or no difference in thought informationists offered (question 3.4, Ta-
the rates at which other members of the team used ble 3). Multiple responses were allowed. By 2006,
additional databases. In 2006, PubMed was still the there were significant increases in perceptions of
most often reported database consulted, with 96% benefits, including ‘‘added thoroughness,’’ ‘‘provid-
(81/84) responding that they used it when questions ing expertise in finding information,’’ and ‘‘help
that could be answered by the literature arose. finding additional information.’’ The one option
To explore whether informationists facilitated or where an informationist was not generally seen as
improved access to information (study question 1c), providing a benefit, ‘‘helping find information for
respondents were asked to rate the frequency with patients,’’ is a function normally reserved for the
which 3 factors caused unsuccessful information nurse educator on NIH clinical research teams and
seeking (question 2.5, Table 2). In 2006, ‘‘lack of time’’ therefore was not unexpected.
(P50.02), ‘‘difficulty finding information’’ (P50.01),
To answer the last study question about requisite
and ‘‘insufficient training’’ (P50.01) were all signifi-
training for informationists, the survey asked respon-
cantly lesser issues than they had been in 2004.
dents to rank six competencies that informationists
Consistent with these findings, respondents felt
should have to be effective team members (question
‘‘more satisfied’’ (P50.01) with their ability to obtain
3.5, Table 3). In both years, the first choice was
answers in 2006 than they had in 2004 (question 2.6,
consistent: ‘‘expertise searching information sources
Table 2), suggesting that informationists had in-
relevant to my clinical/research area.’’ Interestingly,
creased the confidence of clinical research teams in
by 2006 ‘‘specific knowledge of my clinical/research
adequately researching the literature (study question
1d). area’’ increased from the fourth to the second most
important competency.
Attitudes about the informationist program
DISCUSSION
Several survey questions examined researchers’ per-
ceptions and acceptance of the informationist pro- Findings from the survey responses supplied full or
gram. To explore the informationists’ role in clinical partial answers to the study questions and provided
decision making (study question 1e), respondents specific information about how informationists affect-
were asked in which team tasks and activities their ed their teams and how the informationist program
informationists participated (question 3.1, Table 3). has matured over time. Generally, findings indicated
There was a significant increase over time in the levels that the presence of an informationist in the clinical
at which informationists participated in team activi- research environment did help PIs or co-PIs and

152 J Med Libr Assoc 98(2) April 2010

 The informationist

Table 3
Perceptions of informationist contributions and competencies
2004 (n=74) 2006 (n=84) x2 df P

3.1. What does your informationist do?* (multiple responses accepted)

Going on rounds 23% (7) 71% (41) 28.802 1 ,0.0001
Conducting comprehensive searches 53% (17) 81% (47) 17.756 1 ,0.0001
Helping with searches 53% (17) 95% (55) 28.654 1 ,0.0001
Screening and summarizing 20% (6) 43% (25) 11.697 1 ,0.0006
Helping organize my files 7% (2) 31% (18) ,0.0005{
Preparing manuscripts 7% (2) 34% (20) ,0.0001{
Evaluating literature critically 13% (4) 50% (29) ,0.0001{
Other 30% (9) 38% (22) 4.909 1 0.0267
3.2. Would you use the service again?*
Yes 80% (24/30) 100% (58/58) 21.128 1 ,0.0001
Don’t know 6% (6) — —
No — — — —
3.3. Would you recommend the service to others?*
Yes 87% (26/30) 97% (56/58) 15.669 1 ,0.0001
Don’t know 13% (4) — —
No — — 3% (2)
3.4. What are the benefits of an informationist service? (multiple responses accepted)
Providing added thoroughness 76% (56) 92% (77) 7.553 1 0.006
Providing expertise in available databases 80% (59) 100% (84) ,0.0001{
Providing expertise in finding information 80% (59) 94% (79) 7.295 1 0.0069
Saving time 66% (49) 96% (81) ,0.0001{
Helping to find additional information 62% (46) 96% (81) ,0.0001{
Reducing workload burden 58% (43) 89% (75) 20.226 1 ,0.0001
Helping to find information for patients 43% (32) 57% (48) 3.041 1 (ns)
Other 4% (3) 13% (11) 0.0533{
3.5. What are the most important competencies an informationist should have? (top 4 rankings shown)
Expertise searching information sources relevant to my clinical/ 1st 1st
research area
Ability to critically evaluate articles 3rd 3rd
Expertise in evidence-based medicine searches 2nd 4th
Specific knowledge of my clinical/research area 4th 2nd

* Questions asked only of team members who reported working with an informationist (n530 in 2004 and n558 in 2006).
{ Fisher’s exact test used instead of chi-square (see ‘‘Methods’’).

fellows effectively utilize both the growing number on information-related tasks actually increased by
and increasingly complex biomedical resources. 37%. By 2006, survey respondents reported they were
significantly more likely to pursue answers to
Improved access and increased information seeking questions and they were more satisfied with their
frequency and confidence ability to find needed information, findings sugges-
tive of a positive effect over time from having an
Over the period between baseline and follow-up informationist as a team member.
surveys, clinical researchers were more likely to
pursue answers to their questions. Although time Increased range of information sources
has been cited frequently as a leading obstacle for
clinicians seeking information [6, 9], this study Informationists also appear to have had a positive
indicated time became less of an obstacle. This is effect on the range of information services and
even more notable considering that busy PIs and co- resources consulted by PIs or co-PIs and fellows in
PIs made up a higher percentage of respondents in response to an information need. In 2004, most survey
2006 than in 2004. As was demonstrated in both the respondents relied solely on PubMed as a source of
2004 and 2006 survey results, it was generally the lot information, but in 2006, the range and frequency
of fellows to spend a large percentage of their time with which other databases were consulted by PIs or
engaged in information seeking to find answers to co-PIs and fellows rose significantly. This change
clinical questions. However, after 2 or more years cannot be attributed to increased availability or
with an informationist, PIs and co-PIs increased the classroom training, because all the listed resources
likelihood that they would seek answers themselves had been available to NIH staff for several years prior
when questions arose. In addition, information was to the first survey and because classroom training and
viewed as easier to find despite both the constantly one-on-one tutorials in the major end-user databases
increasing numbers of journals, articles, and genetic including GenBank were offered many times prior to
sequences available in online databases, as well as the the first survey. For example, Web of Science had
periodic changes in search features and interface been available at NIH since 1998, 6 years prior to the
design. These findings were especially interesting start of the study, with numerous training sessions
because the survey responses also showed time spent held yearly, yet the data show use of Web of Science

J Med Libr Assoc 98(2) April 2010 153

Grefsheim et al.

by the groups in this study doubled (54% reported recommend one to others. The increased use and
using it in 2006, compared with 27% in 2004) over the recognition of potential benefits no doubt fed each
2-year period. Increased use of the Cochrane database other, the result of the long-term relationships that
(45% of group members reported using in 2006, informationists embedded in teams were able to
compared to just 19% in 2004) is even more likely to build.
be attributable to the informationists, because no One of the more intriguing findings from the
formal classes in this resource had been offered at the surveys was that for a large number of respondents,
NIH Library prior to or during the study period. informationists were engaged in ‘‘other’’ team activ-
Increased use of genetics databases is noteworthy but ities beyond those anticipated. Although the survey
is probably more a reflection of the increasing itself did not provide insight into what these ‘‘other’’
importance of molecular biology and genetics to activities might be, the authors’ experience with the
clinical researchers generally than of the influence of program indicated that these activities included such
informationists. things as creating large citation databases, conducting
bibliometric analyses of grant-funded publications,
Improved clinical and research decision making developing web pages and wikis to facilitate commu-
nication within and outside the teams, compiling and
Whether an informationist improved the clinical or indexing a video database demonstrating movement
research decision making and practice of team disorders, and facilitating use of protocol authoring
members was not clear. What was evident was that, software [26]. Informationists’ adoption of more of
over time, the vast majority of investigators thought these duties with their groups supported the finding
their informationist added needed expertise, found of the recent systematic review [17] that, with
information that they otherwise would not have maturity, clinical informationist programs evolve to
found, added thoroughness, saved them time, and support more technical and informatics needs of
reduced their workload burden. To the extent that clinicians and researchers.
these perceived benefits improved decision making,
informationists had an effect. Study limitations

Education and experience Respondents were volunteers and, therefore, were not
necessarily representative of the groups of which they
While expertise in searching information sources were members. Also, survey methodology addresses
relevant to the team was consistently the informa- people’s perceptions rather than actual events. Thus,
tionists’ most valued knowledge or skill, by 2006, the while most of the study questions related directly to
importance of the informationists’ subject knowledge the survey questions, some questions were answered
also was recognized. This finding suggests that teams only by inference. For example, whether having an
initially adopted informationists because they wanted informationist on the team improved decision making
better access to the literature. Over time, however, the could only be inferred from respondents’ perceptions
informationist’s subject knowledge was recognized as of the informationist. This issue requires further
necessary if all the perceived benefits were to be study. In addition, conducting the two iterations of
achieved. While training informationists in science the survey in different ways—by web and by
and medicine had always been an essential compo- telephone—might have influenced responses. Fur-
nent of the NIH informationist program [23], for ther, the larger percentage of PIs and co-PIs respond-
many other informationist programs, it was merely ing to the second survey might have influenced
desirable [24, 25]. The current study showed that results. The lack of a control group against which to
clinical research teams themselves viewed it as key to compare the findings limited the ability to attribute
the informationists’ preparation. changes in information behavior solely to the infor-
mationists’ presence on the teams. For instance, while
Program maturation the number of electronic resources available to
researchers at NIH did not increase appreciably
Over time, the evolution and uptake of information- during this period, the size, and therefore usefulness,
ists’ services was apparent (question 3.1, Table 3). of databases such as GenBank did change and might
While NIH informationists reached the entire team have influenced their use. Therefore, while the
when they made presentations or participated in presence of informationists on NIH clinical research
rounds, they also worked extensively with individual teams might be statistically associated with changes in
team members. By 2006, more individual team the information behavior of clinical research teams
members, primarily PIs or co-PIs and fellows, were over time, it was not necessarily causal.
working with the informationist than had in 2004,
showing that informationists had achieved greater CONCLUSION
penetration among MDs and PhDs on the team.
Researchers also reported valuing several informa- Given the limitations of the surveys, the conclusions
tionist program features significantly more in 2006 drawn from the findings were conservative. However,
than in 2004. In addition, nearly all respondents over time, informationists on NIH clinical research
reported they would use an informationist again and teams did appear to make a difference in scientists’

154 J Med Libr Assoc 98(2) April 2010

 The informationist

information behaviors, particularly PIs and co-PIs, the 4. Covell DG, Uman GC, Manning PR. Information needs in
people with whom they worked most. For example, office practice: are they being met? Ann Intern Med. 1985
there were significant differences in information Oct;103(4):596–9.
behaviors between 2004 respondents and 2006 re- 5. Osheroff JA, Forsythe DE, Buchanan BG, Bankowitz RA,
Blumenfeld BH, Miller RA. Physicians’ information needs:
spondents in pursuing answers to questions. analysis of questions posed during clinical teaching. Ann
Further, the study indicated that informationists are Intern Med. 1991 Apr 1;114(7):576–81.
made, not born. It not only took time for them to 6. Ely JW, Osheroff JA, Ebell MH, Chambliss ML,
become accepted, contributing members of the clinical Vinson DC, Stevermer JJ, Pifer EA. Obstacles to answering
care or research team, but their education and doctors’ questions about patient care with evidence:
experience in the subject areas and settings in which qualitative study. BMJ. 2002 Mar 23;324(7339):710.
they practice as informationists contributed to their 7. Gorman PN, Ash J, Wykoff L. Can primary care
involvement in team activities. Acknowledging this, it physicians’ questions be answered using the medical
was still their expertise as information scientists that journal literature? Bull Med Libr Assoc. 1994 Apr;82(2):
140–6.
was most highly valued. 8. Westbrook JI, Coiera EW, Gosling AS. Do online
The findings lead to the conclusion that any library information retrieval systems help experienced clinicians
considering establishing its own informationist-type answer clinical questions? J Am Med Inform Assoc. 2005
program should be prepared to invest staff time and May–Jun;12(3):315–21.
money to maintain the information science expertise 9. Coumou HCH, Meijman FJ. How do primary care
of their librarians, as well as develop their subject physicians seek answers to clinical questions? a literature
matter knowledge in the specialty areas of the clinical review. J Med Libr Assoc. 2006 Jan;94(1):55–60.
groups they support. Further, given the growing 10. Giuse NB, Koonce TY, Jerome RN, Cahall M, Sathe NA,
importance of molecular biology and bioinformatics Williams A. Evolution of a mature clinical informationist
model. J Am Med Inform Assoc. 2005 May–Jun;12(3):249–55.
to clinical researchers, shown by the increase in 11. Yarfitz S, Ketchell DS. A library-based bioinformatics
reliance on the key databases in the field, as the services program. Bull Med Libr Assoc. 2000 Jan;88(1):
informationist’s relationship with the clinical team 36–48.
develops, they may need to add bioinformatics skills 12. Chattopadhyay A, Tannery NH, Silverman DAL,
to their armamentarium. Bergen P, Epstein BA. Design and implementation of a
This study has attempted to demonstrate the library-based information service in molecular biology and
impact over time that informationists can have on genetics at the University of Pittsburgh. J Med Libr Assoc.
the groups with whom they work. Future research 2006 Jul;94(3):307–13, E192.
should focus on gaining a deeper understanding of 13. Greenhalgh T, Hughes J, Humphrey C, Rogers S,
Swinglehurst D, Martin P. A comparative case study of
the informationist on clinical and/or research two models of a clinical informaticist service. BMJ. 2002 Mar
teams, particularly the larger effects on health care 2;324(7336):524–9.
quality and health economics. A controlled com- 14. Sladek RM, Pinnock C, Phillips PA. The informationist:
parison of health teams with and without informa- a prospective uncontrolled study. Int J Qual Health Care.
tionist members would contribute to this under- 2004 Dec;16(6):509–15.
standing. Additionally, qualitative studies, using 15. Verhoeven AA, Schuling J. Effect of an evidence-based
interviews and observation, for example, could answering service on GPs and their patients: a pilot study.
provide more details on why researchers choose to Health Info Libr J. 2004 Sep;21 suppl 2:27–35.
work with informationists and what they value 16. Swain R, Oliver KB, Rankin JA, Bonander J, Loonsk JW.
Bioterrrorism alert: reference and literature support for the
about them. As the informationist role develops in CDC’s Director’s Emergency Operations Center (DEOC)
health care settings nationally and internationally, it and investigative field teams. Ref Serv Rev. 2004;32(1):
should be informed and shaped through both local 74–82.
and multisite studies. 17. Rankin JA, Grefsheim SF, Canto CC. The emerging
informationist specialty: a systematic review of the litera-
ACKNOWLEDGMENTS ture. J Med Libr Assoc. 2008 Jul;96(3):194–206. DOI: 10.3163/
1536-5050.96.3.005.
The authors thank Douglas Joubert, NIH Library, and 18. Rogers EM. Diffusion of innovations. 5th ed. New York,
Jason Broadway, National Library of Medicine, for NY: Free Press; 2003.
19. Mulvaney SA, Bickman L, Giuse NB, Lambert EW,
help organizing and analyzing data from the surveys. Sathe NA, Jerome RN. A randomized effectiveness trial of a
clinical informatics consult service: impact on evidence-
REFERENCES based decision-making and knowledge implementation.
J Am Med Inform Assoc. 2008 Mar–Apr;15(2):203–11.
1. Davidoff F, Florance V. The informationist: a new health 20. Rosenbloom ST, Giuse NB, Jerome RN, Blackford JU.
profession? Ann Intern Med. 2000 Jun 20;132(12):996–8. Providing evidence-based answers to complex clinical
2. Institute of Medicine. Crossing the quality chasm: a new questions: evaluating the consistency of article selection.
health system for the 21st century. Washington, DC: Acad Med. 2005 Jan;80(1):109–14.
National Academies Press; 2001. 21. Schacher LF. Clinical librarianship: its value in medical
3. Whitmore SC, Grefsheim SF, Rankin JA. Informationist care. Ann Intern Med. 2001 Dec 4;134(8):717–20.
programme in support of biomedical research: a pro- 22. Grefsheim SF, Rankin JA. Information needs and
gramme description and preliminary findings of an evalu- information seeking in a biomedical research setting: a
ation. Health Info Libr J. 2008;25(2):135–41. study of scientists and science administrators. J Med Libr

J Med Libr Assoc 98(2) April 2010 155

Grefsheim et al.

Assoc. 2007 Oct;95(4):426–34. DOI: 10.3163/1536-5050.95.4. .nih.gov, Chief, Information and Education Services
426. Branch, Division of Library Services/NIH Library;
23. Robison RR. Informationist education. Med Ref Serv Q. National Institutes of Health, 10 Center Drive, MSC
2008 Fall;27(3):339–47.
1150, Bethesda, MD 20892-1150; Barbara A. Rapp,
24. Cataldo TT, Tennant MR, Sherwill-Navarro P, Jesano R.
Subject specialization in a liaison librarian program. J Med PhD, MLS, rappb@nlm.nih.gov, Office of Health
Libr Assoc. 2006 Oct;94(4):446–8. Information Programs Development, National Li-
25. Fikar CR, Corral OL. Non-librarian health professionals brary of Medicine, National Institutes of Health,
becoming librarians and information specialists: results of an Bethesda, MD 20894-3811; Jocelyn A. Rankin, PhD,
Internet survey. Bull Med Libr Assoc. 2001 Jan;89(1):59–67. MLn, AHIP, FMLA, jrankin@cdc.gov, Chief, CDC
26. Robison RR, Ryan ME, Cooper ID. Inquiring informa- Information Center, Centers for Disease Control and
tionists: a qualitative exploration of our role. Evid Based Libr Prevention, Atlanta, GA 30333; Rex R. Robison,
Info Pract [Internet]. 2009;4(1) [cited 23 Apr 2009]. ,https:// PhD, MLS, robisonr@mail.nih.gov, Informationist,
ejournals.library.ualberta.ca/index.php/EBLIP/article/view
/4192/.. Division of Library Services/NIH Library; Candace
C. Canto, MA, MLS,{ cantoc@gmail.com, Special
AUTHORS’ AFFILIATIONS Assistant to the Director, Division of Library Servic-
es; National Institutes of Health, Bethesda, MD
Suzanne F. Grefsheim, MEd, MSLS, FMLA, grefshes@ 20892-1150
nih.gov, Director, Division of Library Services;
Susan C. Whitmore, MS, MSLS, whitmors@mail Received June 2009; accepted September 2009

{ Retired.

156 J Med Libr Assoc 98(2) April 2010