Ij

This material is protected by U.S. copyright law. Unauthorized reproduction is prohibited. To purchase quantity reprints,
please e-mail reprints@ons.org or to request permission to reproduce multiple copies, please e-mail pubpermissions@ons.org.
Oncology Nursing Society Year 2004
Research Priorities Survey
Ann M. Berger, PhD, RN, AOCN®, Donna L. Berry, PhD, RN, AOCN®,
Kimberly A. Christopher, PhD, RN, OCN®, Amanda L. Greene, PhD, MPH, FNP, RN,
Sally Maliski, PhD, RN, MS, Karen K. Swenson, PhD(c), RN, MS, AOCN®,
Gail Mallory, PhD, RN, CNAA, and Dan R. Hoyt, PhD
Key Points . . .
Purpose/Objectives: To determine the Oncology Nursing Society
(ONS) research priorities for 2005­2008 for oncology nursing across
the entire scope of cancer care, including prevention, detection, treat-
Online survey response rate (15%) was lower than for the 2000
ment, survivorship, and palliative care.
mailed paper surveys, but the online method saved time and
Design: Descriptive, cross-sectional survey.
money.
Sample: Stratified into two groups: random sample of general mem-
The top 20 research priorities identified by the general
bership (N = 2,205; responses = 287, or 13%) and all ONS active members
membership included six topics that moved up since 2000:
in the United States with doctoral degrees (N = 627, responses = 144, or
participation in decision making about treatment, evidence-
23%); overall response rate was 15%.
Methods: The 2000 survey was revised and updated. Postcards were
based practice, nurses as advocates, curative treatment/care,
mailed to the original sample (N = 1,605) prior to the launch of the online
cognitive impairment/mental status changes, and late effects
survey, inviting participation via an online or paper-and-pencil survey. An
of treatment. Two topics were new to the 2004 survey: tobacco
e-mail announcement of the survey was launched one week later, followed
use and exposure and initial cancer diagnosis.
by reminders the following week. Because of low response rates, a second
The doctorally prepared sample's top 20 included 10 topics that
sample (N = 600) was selected and contacted.
were not included in the top 20 for the general membership.
Main Research Variables: 117 topic questions divided into seven
categories. Several items were new or reworded.
Despite constant flux, fundamental cancer care topics remain
Findings: The top 20 research priorities included 12 of the top 20
top priority items.
items found in the 2000 survey; 8 topics were new to the top 20. Priority
topics were distributed across six of seven categories. When general
membership results were compared to the doctoral sample, 10 topics
were among the top 20 for both groups. Nine topics were top priorities
rell, Nail, Benedict, & Haberman, 1991; Ropka et al., 2002;
in the 2000 (researcher) and 2004 (doctorally prepared) surveys.
Stetz, Haberman, Holcombe, & Jones, 1995).
Conclusions: Response rates to the electronic survey were lower
In 2003, ONS began using Internet survey methods as a strategy
than for previous paper-and-pencil surveys, but an adequate response
for gathering information from various groups of members. Us-
was obtained. Rank order of mean importance ratings was determined
ing Internet survey methods to assess membershipwide research
by narrow differences in scores. The general membership and doctorally
priorities is unique to ONS among nursing specialty organizations.
prepared samples showed similarities as well as differences in results.
Online surveying is a more cost-effective administration method
Implications for Nursing: The 2004 survey results will inform the
2005 research agenda and assist the ONS Foundation and other funding
organizations in distributing research funds.
Ann M. Berger, PhD, RN, AOCN  ®, is an associate professor and ad-
vanced practice nurse in the College of Nursing at the University of Ne-
braska Medical Center in Omaha; Donna L. Berry, PhD, RN, AOCN  ®,
is an associate professor in Biobehavioral Nursing and Health Systems
C
are of people with cancer is in constant flux, with
at the University of Washington in Seattle; Kimberly A. Christopher,
dramatic changes occurring in areas such as cancer
PhD, RN, OCN  ®, is an assistant professor in the College of Nursing
health behaviors, decision making, delivery systems,
at the University of Massachusetts­Dartmouth in North Dartmouth;
and symptom management. In response to this dynamic situ-
Amanda L. Greene, PhD, MPH, FNP, RN, is a senior researcher at
ation, the Oncology Nursing Society (ONS) has established
NOVA Research Company in Bethesda, MD; Sally Maliski, PhD, RN,
a strong track record of surveying its broad membership to
MS, is an assistant professor in the School of Nursing at the Univer-
assist in determining research priorities for the organization.
sity of California, Los Angeles; Karen K. Swenson, PhD(c), RN, MS,
Beginning in 1981 and conducted about every four years, the
AOCN  ®, is an oncology research manager at Park Nicollet Institute
ONS Research Priorities Survey has sought opinions about
in Minneapolis, MN; Gail Mallory, PhD, RN, CNAA, is director of
research at the Oncology Nursing Society in Pittsburgh, PA; and Dan
research priorities from ONS members to gather a wide
R. Hoyt, PhD, is a professor of sociology and director of the Bureau of
range of experiences with innovations and techniques from
Sociological Research at the University of Nebraska­Lincoln. (Submit-
their respective roles related to the care of people with cancer
ted September 2004. Accepted for publication November 15, 2004.)
(Funkhouser & Grant, 1989; Grant & Stromborg, 1981; Mc-
Guire, Frank-Stromborg, & Varricchio, 1985; Mooney, Fer-
Digital Object Identifier: 10.1188/05.ONF.281-290
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
281
than mailed paper questionnaires and generally is thought to
familiar are you with current research evidence about health
provide at least equivalent results with samples from professional
promotion/disease prevention topics?"). Using Likert-style
organizations (Vehovar, Batagelj, Manfreda, & Zaletel, 2002).
answers, respondents rated their degree of familiarity in each
Although Internet response rates typically have been lower than
of the seven categories as extremely familiar, very familiar,
those for postage-paid, mailed paper surveys, investigators and
moderately familiar, somewhat familiar, and not familiar at
organizations enthusiastically have embraced the benefits of sav-
all. These generic questions for each category were designed
ing time and money (Schonlau, Fricker, & Elliott, 2001).
to elicit respondents' familiarity with the category as a con-
Results of prior surveys have helped direct resources to
textual factor in which to interpret importance ratings.
areas of most importance at those particular times. The ONS
Respondents next rated each of the 117 topic questions
Research Agenda 2003­2005 (ONS, 2003) is one example of
using Likert-style answers as extremely important, very
how the survey results are used. The ONS Foundation has used
important, moderately important, somewhat important, not
survey results to identify funding sources and develop strategic
important at all, or not sure. At the end of the survey, two
initiatives. The purpose of the 2004 survey was to determine the
open-ended questions were posed: (a) Identify the three most
ONS research priorities for 2005­2008 for oncology nursing
important topics in which to conduct new oncology nursing
across the entire scope of cancer care, including prevention,
research and (b) identify other important areas for oncology
detection, treatment, survivorship, and palliative care.
nursing research, including those that are "cutting edge" or
"visionary." Eleven additional questions dealt with the demo-
Methods
graphic and professional characteristics of the respondents.
The entire survey included seven questions related to famil-
ONS Project Team
iarity with the category, seven categories with a total of 117
A 2004 ONS Research Priorities Survey Project Team was
items, 12 demographic questions, and 2 open-ended questions
recruited through the ONS Advanced Nursing Research Spe-
(total = 138) and took about 15 minutes to complete.
cial Interest Group and project team applications. The project
Study Sample
team leader, Ann M. Berger, PhD, RN, AOCN®, worked with
Participants in this survey were recruited from the member-
ONS Director of Research Gail Mallory, PhD, RN, CNAA,
ship of ONS, a national specialty organization for oncology
to select team members who represented a wide variety of
nurses. ONS has more than 30,000 members who work in a
backgrounds and expertise. The principal investigator of the
variety of practice, academic, and research settings. The edu-
ONS Research Agenda Conference, Donna L. Berry, PhD,
RN, AOCN®, was invited to serve on the project team to pro-
cational preparation of members ranges from associate's degree
to doctorate. For this study, the membership was stratified into
vide continuity between the survey and the research agenda.
those who do not hold a doctorate and members who are doc-
A statistician was consulted to assist with data analysis. The
torally prepared. All doctorally prepared members (n = 627)
project team conducted the work associated with constructing
were invited to participate in the survey. Of members who were
and analyzing the results via one team meeting and confer-
not doctorally prepared, 1,605 of the ONS general membership
ence calls. Electronic communication was used to review and
population were selected randomly to participate.
provide feedback throughout the process.
A variety of methods were used to promote response to the
Survey
survey. After two e-mail reminders and two postcards to mem-
bers without e-mail addresses, a total of 301 responses (297
A descriptive, cross-sectional design guided this survey study.
electronic and 4 hard copies) were received. To increase the
After reviewing the 2000 Research Priorities Survey (Ropka et
response rate to a minimum of 400 respondents and to obtain
al., 2002), the project team added new topics, deleted or combined
results with a confidence interval of about plus or minus 5%, an
others, and reorganized the survey format to make it more cur-
additional 600 randomly sampled members who did not hold
rent, respondent friendly, visually attractive, and easy to read. The
doctorates and had e-mail addresses were invited to participate.
final survey consisted of 117 topic questions divided into seven
The final sample size of completed surveys was 431. Of those
categories. New topics were added or previous topics revised
members who accessed the survey Web site, 148 began taking
based on the project team's experience with oncology nursing
the survey and submitted at least the first page of responses but
research topics that have emerged or changed since the last sur-
did not complete the survey. No information is available regard-
vey. In particular, topics related to special populations, ethnicity,
ing why they did not complete the survey. Only completed
and tobacco use recently have gained importance in oncology
surveys (n = 431) are reported in this article.
nursing research. With the exception of the topic on tobacco use,
Only 2% of the general ONS membership has earned a doc-
all additions and modifications were made prior to the original
toral degree; therefore, the survey team decided to stratify based
distribution of the surveys. The tobacco question was added after
on doctorate versus no doctorate. The team hypothesized that
feedback from several members at the 2004 ONS Congress who
those holding a doctorate would have advanced research training
stated that tobacco use had not been addressed adequately in the
and would be more familiar with current research methods and
survey. The tobacco use topic was added to the survey 14 days
findings. Also, doctorally prepared members often are involved
after it went online. This brought the total number of items to 117
actively in conducting oncology nursing research. Without
on all surveys distributed from that date forward, resulting in 224
oversampling, this small group might have been missed with
out of 431 surveys returned containing the tobacco topic.
random sampling. A representative sample of those who are not
A new feature of the 2004 survey was that questions regard-
doctorally prepared was desired because clinicians routinely deal
ing the importance of conducting new research in specific
with problems needing research evidence and use research find-
topics under each category were preceded by a general ques-
ings. Sustaining, associate, and international members of ONS
tion asking respondents to rate their degree of familiarity with
were excluded from the sample. As an incentive to participate,
current research evidence about that category (e.g., "How
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
282
all respondents were eligible for a drawing to win one of three
nonresponse rates, respectively). For the tobacco question that
one-year ONS memberships.
was added later in the survey process, the people who did not
receive that question had it coded as missing. Overall, the level
Data Collection
of item nonresponse was very low, particularly for what typically
To publicize the survey, a postcard that explained the study
is found in self-administered surveys (Dillman, 2000).
purpose was mailed to the entire sample one week prior to the
Analyses were conducted on the full sample treating the
survey launch. The postcard and invitation to participate were
missing and item nonresponses as missing data. Some analyses,
developed by the ONS Marketing Team in consultation with the
such as the basic sample descriptions, were not adjusted for
project team to make the invitation interesting to all ONS mem-
sampling proportions. The general analyses of response patterns
bers. Two options for completing the survey were explained.
and rank orderings were conducted using weighted data to cor-
The postcard listed the Web site for the online survey and gave
rect for potential bias caused by the disproportionate sampling
a toll-free phone number to call to request a paper-and-pencil
of doctorally prepared members. This was accomplished by
survey. The postcard was followed one week later by an e-mail
calculating the weights as described and then applying them in
announcement that was sent to all sampled members who had
SPSS using the program weighting procedure.
an e-mail address listed with ONS (84%). The Web site with
Results
the survey could be accessed directly from the e-mail. After two
weeks, a second reminder postcard was sent to those who did
Response Rates and Demographics
not have e-mail or who requested a paper-and-pencil survey. A
second e-mail reminder was sent to the rest of the sample. When
Of the 2,832 ONS members invited to participate, the
the second sample of 600 was selected, an e-mail announcement
overall response rate to the online survey was 431 (15%).
about the survey was sent. A follow-up e-mail was sent after
Of the doctorally prepared nurses, the response rate was 144
two weeks to encourage participation in the survey. The entire
of 627 members (23%). Among the general membership
sample was notified in the follow-up e-mail that the deadline for
who were invited, the response rate was lower, with 287 of
the survey was extended by about two weeks.
2,205 members (13%) responding to the survey. Only four
To capture the largest response that represents the breadth of
paper-and-pencil surveys were requested in response to the
the ONS membership, several response modes were offered.
postcard mailings. The response rate to the 2004 survey was
Respondents had an option to complete the survey via the Web
lower than the previous 2000 and 1994 surveys, with the
site or through a mailed paper-and-pencil survey. Although
prior response rates to mailed questionnaires being 39% and
many ONS members have access to the Internet, about 22%
36%, respectively (Ropka et al., 2002; Stetz et al., 1995).
have not reported an e-mail address to ONS. In the current sur-
The majority of respondents to the 2004 survey were female
vey, 15.8% (n = 447) of the sample did not have a listed e-mail
(97%) and Caucasian (89%), which reflects the makeup of
address and therefore received a total of two postcard mailings
the ONS membership (see Table 1). However, the percentage
that gave them the option to request that a paper-and-pencil sur-
vey be mailed to them or to go directly to the Web site using an
Table 1. Demographics of 2004 Survey Respondentsa and
available computer. Additionally, those with an e-mail address
Oncology Nursing Society (ONS) Membershipb
had the option of requesting a paper-and-pencil survey.
The online survey was conducted using Zoomerang (Mar-
ONS Membershipb
ketTools, Inc., Mill Valley, CA), a software program that
Survey Respondentsa
(as of April 1, 2004)
facilitates designing and sending surveys and analyzing their
results. An Internet portal was set up to access the survey
Variable
n
%
n
%
online via the ONS Web site. A secure Web site was used
for electronic responses. Confidentiality was ensured by as-
Gender
signment of an identification number to each survey so that
Female
418
97
26,693
97
no individual respondent was identifiable. Respondents were
Male
11
3
965
3
asked to provide their names and addresses if they wanted to
Age (years)
be entered into a drawing to receive the incentive of a one-year
< 30
7
2
1,993
7
30­39
67
16
5,830
20
membership to ONS. Three were selected randomly from the
40­49
161
37
10,101
34
names to receive a free one-year ONS membership. Personal
50­59
163
38
7,026
24
identifying information was not entered into the analysis and
60­69
26
6
1,187
4
was not available to any team member. The survey was con-
> 70
5
1
382
1
ducted over five weeks during April and May 2004.
Ethnicity
The data from the Zoomerang survey were saved into a trans-
African American
12
3
891
3
ferable data format and read into the SPSS® (SPSS Inc., Chi-
Asian
16
4
1,309
5
cago, IL) statistical package for data analysis and interpretation.
Caucasian
381
89
23,577
88
Any item nonresponse because an item was skipped was coded
Hispanic/Latino
6
1
606
2
as missing data and excluded from the analysis. Any "not sure"
Other
16
4
249
1
responses to the Likert-type ranking questions also were coded
Survey respondents: N = 431
as nonresponses. The majority of the questions had less than 1%
a
nonresponse. The highest proportions of nonresponse were on
Categories will not total 431 as a result of nonresponse on selected items.
the special cancer population questions. Of those, the greatest
b
The total N for each response category represents the number of ONS mem-
proportion of nonresponse was for the items on prisoners and
bers who reported their status on ONS membership profiles.
people with gay, lesbian, or bisexual orientation (4.4% and 3.7%
Note. Because of rounding, percentages may not total 100.
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
283
Table 2. Professional Characteristics of Survey Respondents
of survey respondents who were 50 years and older (45%)
and Oncology Nursing Society (ONS) Membership
was higher than the percentage of ONS general membership
who are 50 years and older (29%). This is reflective of the
ONS Membershipb
oversampling procedure that included all of the doctorally
Survey Respondentsa
(as of April 1, 2004)
prepared nurses. Among the doctorally prepared respon-
dents, 61% indicated that they were 50 or older. In contrast,
Variable
n
%
n
%
37% of the nondoctorally prepared respondents were in the
50 or older age group.
Highest degree
in nursing
Description of Survey Respondents
Diploma
38
9
3,892
14
Professional characteristics, years in nursing and oncology
Associate's
54
13
7,130
26
Bachelor's
113
27
10,909
40
nursing, primary functional area, practice setting, and em-
Master's
108
26
4,884
18
ployment status of the respondents as compared to the ONS
Doctorate
110
26
405
1
general membership (as of April 1, 2004) are shown in Table
None
­
­
195
1
2. Survey respondents were more likely than the ONS general
Highest non-
membership to have a master's degree (26% versus 18%) or
nursing degree
a doctorate degree (26% versus 1%) as the highest degree in
Associate's
40
9
1,103
4
nursing. They also were more likely to have a greater num-
Bachelor's
57
13
3,191
10
ber of years of experience in oncology nursing, with 47% of
Master's
32
8
1,380
4
survey respondents compared to 26% of the ONS general
Doctorate
43
10
249
1
membership having 16 or more years in oncology nursing. A
None
252
59
3,602
81
Years in oncology
greater proportion of survey respondents listed research (21%
nursing
versus 8%) and education (21% versus 7%) as their primary
<4
34
8
7,572
26
functional areas as compared to the ONS general membership.
4­10
99
24
8,755
30
Survey respondents also were slightly more likely to be work-
11­15
89
21
5,511
19
ing full-time than the ONS general membership (87% versus
16­20
81
19
4,253
15
79%). The differences between the survey respondents and
21+
115
28
3,237
11
the ONS general membership are explained by the intentional
Years in nursing
oversampling of doctorally prepared nurses.
<4
6
1
4,323
15
4­10
39
9
5,601
19
Top 20 Research Priorities
11­15
51
12
3,824
13
16­20
59
14
4,281
15
Mean importance ratings were calculated for each topic
21+
272
64
11,263
38
and then adjusted to remove the effects of oversampling the
Primary
doctorally prepared group. The adjustment was accomplished
functional area
by assigning sampling weights that represented the inverse of
Patient care
191
45
19,435
71
the ratio of the proportion of nondoctoral respondents in the
Research
91
21
2,059
8
sample with the known proportion of nondoctoral members of
Education
88
21
1,932
7
the organization (determined by membership data). The use of
Administration
34
8
2,495
9
this sampling weight ensures that responses of the oversampled
Other
23
5
1,462
5
Practice setting
doctoral group will be proportional to their actual representa-
Hospital
107
25
12,607
46
tion in the organization. Topics were listed in rank order from
Outpatient and
102
24
4,203
15
most important to least important. Data were reviewed, and
ambulatory care
the team determined that no logical cut point existed in mean
Physician office
29
7
4,524
16
importance ratings. However, a decision was made to report
Other
182
43
6,137
22
the top 20 ratings because this was consistent with the 2000
Current employ-
Research Priorities Survey (Ropka et al., 2002) and would fa-
ment status
cilitate comparisons (see Table 3). All 20 mean importance rat-
Full-time
358
87
22,465
79
ings demonstrated high importance; rating scores ranges from
Part-time
47
11
5,119
18
1.52­1.70 on a scale of 1 (extremely important) to 5 (not at all
Not working
8
2
690
2
important). Some mean importance ratings were the same, yet
Survey respondents: N = 431
the rank order was different. This is because the team reported
a
Categories will not total 431 as a result of nonresponse on selected items.
the results rounded to within two decimal points of accuracy.
b
The total N for each response category represents the number of ONS mem-
The rank order for items that appeared to be tied was based on
bers who reported their status on the ONS membership profiles.
the ranking from the nonrounded means. Clearly, any distinc-
Note. Because of rounding, percentages may not total 100.
tion among items that are equivalent out to two decimal points
is very minor and nonsignificant.
compared to the 2000 general membership sample (Ropka
Comparison Between the 2000 and 2004 Research
et al., 2002). When comparing the rank order of the 2004
Priorities
total sample with the previous surveys, the team found fewer
Table 3 also compares the rank order of the top 20 priority
changes in priorities between 2000 and 2004 (12 items the
topics identified by the general membership sample in 2004
same) than between the prior 1994 and 2000 surveys (9
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
284
Table 3. Top 20 Research Priorities Ranked by Mean
survey by the total sample. These topics were neutrope-
Importance Ratings for the Total Membership Sample,
nia/immunosuppression, oncologic emergencies, suffering,
Adjusted for Doctorate Group Oversamplinga,
anorexia, access to cancer care, depression, coping/adapta-
With Comparisons to the 2000 Survey
tion, nurse workforce issues, and hope. Mean importance
ratings for these items fell below the cut point of 1.70 in 2004
­
X
2000 Survey
(range = 1.70­1.98).
Rank
Importance
Rank
Rank Order of Mean Importance Ratings
Ratingb (SD)
Order
Topic
Order
Table 4 displays each category with its respective topics
Quality of life
1.52 (0.718)
2
1
listed by mean importance rating and rank order among all
Participation in decision
1.54 (0.701)
18
2
topics as determined by the total sample and adjusted for
making about treatment in
doctoral group oversampling. All questionnaire categories
advanced diseasec
demonstrated some variability in topic mean importance
Patient/family educationc
1.55 (0.742)
19
3
ratings; however, no topic had a mean rating higher than
Participation in decision
1.58 (0.701)
43
4
2.94.
making about treatmentc
Pain
1.59 (0.746)
1
5
A bold-faced topic entry in Table 4 indicates a top 20 rank-
Tobacco use and exposured
1.60 (0.848)
­
6
ing. The top 20 priorities were distributed among six of the
Screening/early detection of
1.60 (0.746)
3
7
seven questionnaire categories. Only the special cancer popu-
cancer
lations category had no topic in the top 20. The number of top
Prevention of cancer/cancer
1.61 (0.754)
4
8
20 topics varied among the six categories. Cancer symptom
risk reduction
management had three priority topics: pain, fatigue/lack of
Palliative care
1.62 (0.727)
17
9
energy, and cognitive impairment/mental status changes.
Evidence-based practice
1.62 (0.810)
78
10
Behavioral/psychosocial aspects of cancer had one priority
Nurses as advocatesd
1.65 (0.773)
­
11
topic: quality of life. Health services had two priority topics:
Fatigue/lack of energy
1.66 (0.773)
9
12
evidence-based practice and patient outcomes of cancer care,
Cancer recurrence
1.67 (0.704)
20
13
Curative treatment/carec
1.67 (0.733)
37
14
a revised item in 2004 that was changed from "outcomes of
1.67 (0.743)
8
Patient outcomes of cancer
15
cancer care" in the 2000 survey. Cancer continuum of care
carec
had eight of the top 20 topics: screening/early detection of
1.67 (0.750)
70
Cognitive impairment/mental
16
cancer, prevention of cancer/cancer risk reduction, palliative
status changesc
care, cancer recurrence, curative treatment/care, late effects
Late effects of treatmentc
1.68 (0.722)
24
17
of treatment, hospice/end of life, and initial cancer diagnosis.
1.69 (0.790)
6
Hospice/end of life
18
Health promotion/disease prevention behaviors had one prior-
Initial cancer diagnosisd
1.69 (0.764)
­
19
ity topic: tobacco use and exposure, a new item in 2004. This
1.70 (0.752)
10
Ethical issues
20
item was added to the survey 14 days after it went online: 207
of the total number of respondents had completed surveys
a
Adjusted by weighting to correct for oversampling of doctorally prepared
nurses
before this question was included. Mean scores were tabulated
for the 224 responses to that question, with the remaining por-
b
Scores ranged from 1 (extremely important) to 5 (not at all important).
tion of the total 431 responses treated as missing data. Com-
c
Item wording was not identical to the 2000 survey.
munication and decision making had five of the top priority
d
New question, not asked in the 2000 survey
topics: participation in decision making about treatment in
advanced disease, patient/family education, participation in
decision making about treatment, nurses as advocates, and
items the same). Twelve topics were ranked among the top
ethical issues.
20 priorities in both the 2004 and 2000 surveys: quality of
life, participation in decision making about treatment in
Doctoral Sample Rankings Versus General
advanced disease, patient/family education, pain, screen-
Membership Rankings
ing/early detection of cancer, prevention of cancer/cancer
Table 5 displays the 2004 top 20 research priority topics
risk reduction, palliative care, fatigue/lack of energy, cancer
for the doctorally prepared sample compared to the top 20
recurrence, patient outcomes of cancer care, hospice/end of
priority topics for the adjusted general membership sample.
life, and ethical issues. Although the 12 items remained in
Ten topics were among the 20 priorities for both groups:
the top 20 priorities in both surveys, the rank order changed
cognitive impairment/mental status changes, evidence-based
for each topic.
practice, late effects of treatment, participation in decision
Six topics not ranked among the top 20 in 2000 were
making about treatment in advanced disease, patient outcomes
included among the top 20 priority topics in the 2004 total
of cancer care, cancer recurrence, palliative care, participa-
sample. These were evidence-based practice, nurses as advo-
tion in decision making about treatment, hospice/end of life,
cates, curative treatment/care, patient outcomes of cancer care,
and fatigue/lack of energy. Four of these topics were from
cognitive impairment/mental status changes, and late effects
the cancer continuum of care category: late effects of treat-
of treatment. Two topics that were new to the 2004 survey also
ment, cancer recurrence, palliative care, and hospice/end of
were ranked among the top 20: tobacco use and exposure and
life. Three additional categories each had two topics: Cancer
initial cancer diagnosis.
symptom management had late effects and fatigue/lack of en-
Nine topics that previously were among the top priori-
ergy, health services had evidence-based practice and patient
ties in the 2000 survey (Ropka et al., 2002) no longer were
outcomes of cancer care, and communication and decision
ranked among the top 20 research priorities in the 2004
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
285
Table 4. Rank Order of Mean Importance Ratings Listed Within Questionnaire Categories, Adjusted for Doctorate Group
Oversamplinga
­
­
X
Rank Order
X
Rank Order
Rank in
Category
Importance
Among All
Rank in
Category
Importance
Among All
Ratingb (SD)
Ratingb (SD)
Category
and Topic
Topics
Topics
Category
and Topic
Cancer symptom management
Behavioral/psychosocial aspects of cancer (continued)
1.70 (0.797)
21
2
1
1.59 (0.746)
Pain
Depression
5
1.77 (0.842)
35
3
2
1.66 (0.773)
Fatigue/lack of energy
Suffering
12
1.77 (0.801)
36
4
3
1.67 (0.750)
Cognitive impair-
Coping/adaptation
16
1.78 (0.808)
39
5
Family functioningc
ment/mental status
1.79 (0.852)
42
6
changesc
Hope
1.79 (0.896)
43
7
4
1.72 (0.829)
Spiritual well-being
Clustering of symp-
23
1.83 (0.866)
48
8
tomsd
Caregiver role
1.89 (0.812)
61
9
5
1.74 (0.768)
Self-management/
Anorexia
27
Neuropathiesc
self-efficacyd
6
1.75 (0.792)
31
1.89 (0.845)
62
10
7
1.75 (0.835)
Oncologic emergencies
Grief
32
2.01 (0.841)
79
11
8
1.80 (0.802)
Functional status
Psychoneuroimmu-
46
changesd
nologyd
2.01 (0.896)
80
12
9
1.83 (0.813)
Dyspnea/shortness of
Anxiety
49
2.02 (0.814)
83
13
breath
Counseling
2.04 (0.855)
89
14
Appetite/taste changesc
10
1.86 (0.797)
Social support
51
2.09 (0.904)
95
15
11
1.87 (0.825)
Body image/sexuality
Cardiac changes (e.g.,
55
rhythm, cardiac
failure)c
Health services
12
1.87 (0.849)
Stomatitis/mucositis
56
13
1.91 (0.880)
Extravasations
63
1.62 (0.810)
10
1
Evidence-based practice
14
1.92 (0.838)
Sleep disorders/
65
1.67 (0.743)
15
2
Patient outcomes of
insomnia
cancer carec
15
1.92 (0.862)
Fluid and electrolyte
66
1.71 (0.807)
22
3
Health legislation
imbalance
policyc
Anemiad
16
1.93 (0.825)
67
1.73 (0.828)
24
4
Nurse workforce issuesc
17
1.94 (0.879)
Bleeding
69
1.74 (0.802)
26
5
Safety/adverse eventsd
18
1.98 (0.953)
Neutropenia/immuno-
75
1.75 (0.822)
30
6
Standards of care
suppression
1.81 (0.866)
47
7
Access to cancer care
Lymphedemad
19
2.00 (0.859)
78
1.87 (0.940)
54
8
Occupational hazards
20
2.02 (0.830)
Sexual dysfunction
84
for cancer nurses
21
2.02 (0.930)
Nausea/vomiting
85
1.88 (0.809)
58
9
Patient/family resource
22
2.03 (0.840)
Hormone disturbanc-
86
supportd
esc
1.94 (0.954)
68
10
Continuing education/
23
2.03 (0.798)
Altered mobility
87
professional develop-
24
2.06 (0.827)
Osteopenia/osteopo-
90
ment
rosisd
1.96 (0.867)
71
11
Economic outcomes of
25
2.06 (0.903)
Fever
91
cancer cared
Thrombocytopeniad
26
2.06 (0.926)
92
1.99 (0.928)
76
12
Continuous quality
27
2.12 (0.858)
Diarrhea
98
improvement
28
2.14 (0.820)
Dry mouth
100
2.00 (0.912)
77
13
Care delivery settings
29
2.20 (0.908)
Wounds
102
2.02 (0.903)
82
14
Advanced practice
30
2.22 (0.831)
Skin changes/
103
nursing
cutaneous reactions/
2.08 (0.964)
94
15
Information systems
urticariac
in patient care
31
2.28 (0.894)
Weight changes (gain
106
2.14 (0.909)
99
16
Case management/
or loss)
care management
32
2.28 (0.911)
Constipation
107
2.20 (0.926)
101
17
Telehealth/remote
Urinary incontinenced
33
2.29 (0.917)
109
assessment and
34
2.35 (0.894)
Cough
111
interventiond
35
2.56 (0.908)
Alopecia
115
2.26 (1.034)
104
18
Acuity/patient classifi-
cation systems
Behavioral/psychosocial aspects of cancer
(Continued on next page)
Quality of lifee
1
1.52 (0.718)
1
a
Adjusted by weighting to correct for oversampling of doctoral nurses
b
Scores ranged from 1 (extremely important) to 5 (not at all important).
c
Item wording was not identical to the 2000 survey.
d
New question, not asked in the 2000 survey
Note. A bold topic entry indicates a top 20 ranking.
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
286
Table 4. Rank Order of Mean Importance Ratings Listed Within Questionnaire Categories, Adjusted for Doctorate Group
Oversamplinga (Continued)
­
­
X
Rank Order
X
Rank Order
Rank in
Category
Importance
Among All
Rank in
Category
Importance
Among All
Ratingb (SD)
Ratingb (SD)
Category
and Topic
Category
and Topic
Topics
Topics
Cancer continuum of care
Special cancer populations (continued)
7
1.60 (0.746)
1.98 (0.894)
72
1
7
Screening/early
Racial/ethnic/cultural
groupsc
detection of cancer
AIDS/HIV
8
1.61 (0.754)
2.07 (0.879)
93
2
8
Prevention of cancer/
Men
9
2.12 (0.887)
97
cancer risk reduction
Disabled
10
1.62 (0.727)
2.28 (0.909)
105
3
9
Palliative care
Rural/frontier
11
1.67 (0.704)
2.29 (0.975)
108
4
13
Cancer recurrence
Mentally ill
12
1.67 (0.733)
2.33 (0.961)
110
5
14
Curative treatment/
cared
Immigrant
13
2.36 (0.982)
112
Internationald
14
1.68 (0.722)
Late effects of treat-
2.44 (0.970)
113
6
17
mentc
15
Migrant
2.47 (1.018)
114
16
Gay/lesbian/bisexual/
1.69 (0.790)
Hospice/end of life
2.71 (1.094)
116
7
18
transgenderedd
1.69 (0.764)
Initial cancer diag-
8
19
Prisonersd
nosisd
17
2.94 (1.055)
117
1.80 (0.762)
Genetic counseling/
9
45
Communication and decision making
testing/treatment
Survivorshipd
1
Participation in deci-
1.87 (0.890)
1.54 (0.701)
2
10
53
sion making about
1.98 (0.831)
Rehabilitation
11
74
treatment in ad-
2.04 (0.866)
Bereavement care
12
88
vanced diseasec
Health promotion/disease prevention behaviors
2
Patient/family edu-
1.55 (0.742)
3
cationc
1.60 (0.848)
Tobacco use and
1
6
exposured
3
Participation in deci-
1.58 (0.701)
4
1.79 (0.863)
sion making about
Stress management
2
40
treatmentc
1.84 (0.770)
Diet/nutrition
3
50
Nurses as advocatesd
Risk appraisald
4
1.87 (0.837)
1.65 (0.773)
11
4
52
5
1.88 (0.788)
Ethical issues
1.70 (0.752)
20
Exercise/physical
5
57
6
Decisions to participate
1.74 (0.770)
25
activity
2.02 (0.915)
in prevention and/or
Substance abuse
6
81
screeningd
2.12 (0.879)
Sleep/rest
7
96
7
Informed consent
1.75 (0.839)
28
Special cancer populations
8
Advance directives
1.75 (0.826)
29
9
1.76 (0.780)
Compliance/adherence
1.77 (0.786)
34
Children/adolescents
1
33
10
1.78 (0.783)
Decisions to participate
1.78 (0.764)
38
Older adults/elderly
2
37
in research studiesc
1.79 (0.813)
Women
3
41
Consumer educationd
11
1.80 (0.796)
1.89 (0.795)
60
Children of parents
4
51
12
Comprehensive health
1.92 (0.832)
64
with cancer
assessmentd
1.89 (0.880)
Socioeconomically
5
59
13
Complementary and
1.98 (0.870)
73
disadvantaged
Low health literacyc
1.96 (0.922)
alternative therapies
6
70
a
Adjusted by weighting to correct for oversampling of doctoral nurses
b
Scores ranged from 1 (extremely important) to 5 (not at all important).
c
Item wording was not identical to the 2000 survey.
d
New question, not asked in the 2000 survey
Note. A bold topic entry indicates a top 20 ranking.
making had participation in decision making about treatment
the seven categories (all except communication and decision
in advanced disease and participation in decision making
making). Special cancer populations had four topics: older
about treatment.
adults/elderly, socioeconomically disadvantaged, low health
Doctorally prepared respondents ranked an additional 10
literacy, and racial/ethnic/cultural groups. Cancer symptom
topics in the top 20 research priorities that the adjusted general
management had two topics: clustering of symptoms and
membership sample did not rank. These topics were older
functional status changes. The remaining four categories each
adults/elderly, clustering of symptoms, socioeconomically
had one priority topic: Behavioral/psychosocial aspects of
disadvantaged, racial/ethnic/cultural groups, access to cancer
cancer had self-management/self-efficacy, health services had
care, exercise/physical activity, low health literacy, functional
access to cancer care, cancer continuum of care had survivor-
status changes, self-management/self-efficacy, and survivor-
ship, and health promotion/disease prevention behaviors had
ship. These 10 priority topics were dispersed among six of
exercise/physical activity.
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
287
Table 5. Top 20 Research Priorities Ranked by Mean
Doctorally Prepared 2004 Rankings Versus
Importance Ratings for the Doctorate Samplea, With
Researcher 2000 Rankings
Comparisons to the 2004 Survey General Membership
Table 5 also displays the top 20 research priority topics for
Sample Rankings and the 2000 Survey Researcher Top 20
the 2004 doctorally prepared nurses and compares them with
the top 20 priorities of the researchers surveyed in 2000. Nine
Rank Order
topics were in the top 20 priorities for both groups: cognitive
impairment/mental status changes, evidence-based practice,
2004 Survey 2004 Survey 2000 Survey
­
late effects of treatment, socioeconomically disadvantaged,
X
Doctoral
General
Researcher
Importance
Sample
Membership
Sample
patient outcomes of cancer care, palliative care, access to
Ratingb (SD)
Top 20c
Topic
Top 20
Sample
cancer care, hospice/end of life, and fatigue/lack of energy.
Although the nine items remained in the top 20 priorities in
Older adults/
1.53 (0.675)
01
37
­
both surveys, the rank order changed for each topic.
elderly
Cognitive
1.56 (0.777)
02
16
11
Familiarity With Current Research Evidence
impairment/
At the beginning of each section on the 2004 survey, re-
mental status
spondents were asked to rate how familiar they were with
changes
current research evidence about each of the topic categories
Evidence-based
1.57 (0.831)
03
01
10
practice
from 1 (extremely familiar) to 5 (not familiar at all). Results
Late effects of
1.58 (0.680)
19
04
17
presented in rank order in Table 6 show that survey respon-
treatment
dents were most familiar with current research evidence
­
Clustering of
1.58 (0.857)
­
05
23
about cancer symptom management (X = 2.54), followed by
­
symptoms
behavioral/psychosocial aspects of cancer (X = 3.27), com-
­
Socioeconomi-
1.58 (0.733)
14
06
59
munication and decision making (X = 3.28), health services
cally disad-
­
(X = 3.31), health promotion/disease prevention behaviors
vantaged
­
­
(X = 3.34), and cancer continuum of care (X = 3.36). The least
Participation
1.64 (0.839)
­
07
02
familiar category was special cancer populations with a mean
in decision
rating of 3.56, which also was the category that did not include
making about
treatment
any top 20 priority rankings.
in advanced
disease
Discussion
Patient out-
1.65 (0.775)
15
08
04
comes of
Findings from the 2004 Research Priorities Survey have
cancer care
provided important information for use by ONS. Determi-
Cancer
1.67 (0.723)
13
­
09
nation of the research priorities by members will guide the
recurrence
allocation of resources to areas of highest priority at this
Racial/eth-
1.69 (0.835)
72
­
10
time. Many priority topical areas of research are complex and
nic/cultural
require sustained focus. Other topics did not remain among
groups
the top priority areas for research, perhaps as a result of the
Palliative care
1.72 (0.826)
16
11
09
dynamic cancer care delivery environment. Emerging priority
Access to
1.73 (0.696)
12
47
08
areas for research have been identified.
cancer care
The overall response rate to the 2004 survey was 15%.
Participation
1.74 (0.847)
13
­
04
in decision
Although this rate is lower than rates for the 1994 and 2000
making about
surveys, it is consistent with figures from electronic surveys
treatment
(Dillman, 2000; Vehovar et al., 2002). Responses to electronic
Exercise/physi-
1.74 (0.857)
14
­
57
surveys conducted by ONS in 2003 ranged from 5%­60%,
cal activity
with higher rates experienced when the survey topics were
Low health
1.74 (0.877)
15
­
70
aimed toward a particular group (ONS, 2004). This was
literacy
evident in the current survey about research priorities, when
Functional
1.77 (0.842)
16
­
46
the general membership response rate was 13%, whereas the
status
doctorally prepared members' response rate was 23%. The
changes
15% response rate was adequate to make scientifically sound
Hospice/end
1.78 (0.856)
17
18
18
of life
conclusions from the findings, according to sample size meth-
Self-manage-
1.78 (0.840)
18
­
61
odology described by Dillman. Three free ONS membership
ment/self-
renewals were offered as incentives to complete the survey.
efficacy
Additional incentive methods need to be identified to encour-
Survivorship
1.80 (0.857)
19
­
53
age survey responses by ONS members in the future.
Fatigue/lack
1.81 (0.894)
20
12
07
The team was not surprised to learn that survey respondents
of energy
were more likely to have a master's or doctoral degree as well
a
as more years of experience in oncology nursing. Those who
N = 144 for the doctorate sample
were invited to participate in the survey had varying levels
b
Scores ranged from 1 (extremely important) to 5 (not at all important).
of enthusiasm for completing it. As might be expected, those
c
Rankings for the 2000 researcher sample were available only for the top 20
with higher levels of education and more experience in oncol-
responses.
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
288
Table 6. Rank Order of Mean Familiarity Ratings for Topic Categories, Percentage Within Category, Adjusted for Doctorate
Oversamplinga
­
1
2
3
4
5
X
Extremely
Very
Moderately
Somewhat
Not at All
Familiarity
Ratingb
Rank
Topic
Familiar (%)
Familiar (%)
Familiar (%) Familiar (%)
Familiar (%)
1
Cancer symptom management
18
05
2.54
33
29
15
2
Behavioral/psychosocial aspects of cancer
11
3.27
06
16
35
33
3
Communication and decision making
04
13
3.28
17
41
26
4
Health services
12
3.31
03
18
35
32
5
Health promotion/disease prevention behaviors
11
3.34
03
15
40
32
6
Cancer continuum of care
15
3.36
02
18
37
28
7
Special cancer populations
17
3.56
01
11
36
35
a
Adjusted by weighting to correct for oversampling of doctorally prepared nurses
b
Scores ranged from 1 (extremely familiar) to 5 (not familiar at all).
ogy nursing chose to respond more frequently. Of note, the
is unsure why some of the eight topics dropped out of the top
same individuals also may have been more likely to have the
20 priorities among general membership in 2004. Areas such
time to be able to complete the survey while at work on an
as depression, access to cancer care, and nurse workforce is-
office computer; they identified their primary functional roles
sues, to name a few, have not been resolved. Mean importance
as researchers and educators.
ratings of these topics (1.70­1.98) were very close to the cut
The rank order of mean importance ratings was determined
point (1.70). The best explanation for these findings is that
by very narrow differences in scores, with no topic having a
very few points separated the items that were identified as
mean score higher than 2.94. This may reflect that the survey
top priority from those that were not ranked in the top 20. In
design or included items did not differentiate clearly among
contrast, workforce issues, specifically job satisfaction, nurse-
research priorities. However, when items were ranked by the
patient ratios and staffing, and nurse retention, were ranked
mean score, a distinct change in ranking from the 2000 survey
among the top five priorities in a Delphi survey of nursing
to the current one was found. This shift in rankings may be
research priorities conducted by Cohen et al. (2004).
a better indication of priorities than the actual scores. Future
The top 20 priorities for the general membership are distributed
surveys may want to reexamine the survey design and survey
broadly among six of the seven categories of the questionnaire,
items to better differentiate among priorities.
with only special cancer populations having no items listed as
Twelve items ranked in the top 20 priority topics of the total
priority topics. This may be because nurses caring for these popu-
membership sample, adjusted for doctoral oversampling, in
lations were not represented adequately in the survey responses.
2004 that also had appeared in the top 20 in the 2000 survey
These rankings provide direction to ONS, the ONS Foundation,
(Ropka et al., 2002). Rank order changed for these 12 items,
and other funding organizations in regard to areas of research to
with only quality of life, pain, screening/early detection of
support. Some topics might have ranked lower than others because
cancer, and prevention of cancer/cancer risk reduction rank-
of lack of awareness on the part of the survey respondents of the
ing in the top 10 in both surveys. Similarly, the item labeled
need for research in the particular area, rather than because of its
acute and chronic pain was ranked first in a nursing research
lack of importance. Examples of items that received scores of
priorities survey recently completed at a large comprehensive
lower priority are alopecia, body image/sexuality, acuity/patient
cancer center (Cohen, Harle, Woll, Despa, & Munsell, 2004).
classification systems, and several cancer populations.
These findings point out that despite constant flux, fundamen-
In comparison, rank order of mean importance ratings by
tal cancer care topics remain top priority items, in part because
the doctoral sample demonstrated different findings than the
their complexity has not permitted them to be resolved.
general membership results. This is not unexpected, as doc-
Topics that were new to the top 20 listing represent trends in
torally prepared members may have a different awareness of
healthcare practice settings, and the authors recommend that
topics than the general membership, such as research on older
these topics be examined carefully and considered for integra-
adults/elderly and socioeconomically disadvantaged people.
tion when ONS develops its research agenda for 2005­2008
However, 10 topics were included among the top 20 priorities
and strategic plan for 2005­2007. Rank order of items related
for both groups, and these findings provide direction for focus
to participation in decision making about treatment and evi-
areas in the next few years.
dence-based practice rose considerably in ranking compared
The display of the 2004 top 20 priority topics for the doctor-
to the 2000 survey results.
ally prepared members with the top 20 priority topics for the
As multiple treatments have become available for various
researchers surveyed in 2000 includes nine topics that were
cancers that have equal or unproven superior survival benefit,
in both listings, with changes in rank order. These findings
those diagnosed with cancer have been faced with making
demonstrate that much research needs to be done in areas such
decisions about their own treatment. With this has come the
as cognitive impairment/mental status changes and evidence-
need to understand treatment decision making such that oncol-
based practice. The findings also serve as a sign that a larger
ogy nurses can better facilitate informed and shared decision
cohort of investigators are needed to demonstrate programs
making. The results provided several surprises, and the team
of research in high-priority areas to make rapid and sustained
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
289
progress in improving outcomes. Students in higher degree
time to identify areas for future research inquiry to ultimately
programs are urged to consider pursuing research on topics
meet the mission of ONS. Feedback regarding the content of
that are recognized as top priority areas in this survey. The
these responses is beyond the scope of this article but will be
2000 survey of researchers only presented the top 20 research
used by ONS when describing cutting-edge research ideas and
priorities. Therefore, the team could not evaluate to what extent
planning future research priorities surveys.
the new 2004 items had advanced from any previous research-
The 2004 Research Priorities Survey Project Team se-
ers' priority ranking in 2000. Examples of such items are older
lected a design that invited responses from a wide range of
adults/elderly and the socioeconomically disadvantaged.
members who identified themselves as clinicians, educators,
Ropka et al. (2002) concluded their report of the 2000 ONS
and researchers. The ONS core value of inclusiveness was
research priorities by recommending that "Future surveys
incorporated into the survey sampling methods. This sampling
might include the opportunity for respondents to comment on
technique exemplifies the value that ONS places on members'
the degree to which research has been applied successfully to
needs and interests. Survey results that were generated from
clinical problems" (p. 490). The authors of the current article
a broad sampling of the ONS membership provide the best
were concerned that the prevalence of certain clinical problems
representation of all members' interests and perceptions of
was influencing the ranking of these research topics instead of
research priorities. These broad-based survey results can
the respondents' understanding of practice evidence. Therefore,
promote research studies that are driven by issues in clini-
a new feature was added to the survey for 2004, a query for
cal practice. This design also emphasizes the importance of
each category of topics regarding respondents' familiarity with
promoting clinician and researcher partnerships in all phases
current research evidence about that category. Not surprisingly,
of the research process. The goals are to generate new knowl-
the clinician-dominated sample was most familiar with the
edge that addresses clinical issues and problems, use research
category of cancer symptom management, and 3 of the top 20
findings in practice, and develop and use evidence-based
topics belonged to this grouping. This information can be used
guidelines for care of patients with cancer.
to further examine the association between familiarity with
Results from this survey will serve as the "voice of the
practice evidence and perceived priority for conducting new
membership" when ONS's 2005­2008 research agenda is de-
research. Additional analysis of these data is planned.
veloped. The results also will be useful to the ONS Foundation
Respondents frequently answered the two questions at the
and other sources of funding for oncology nursing research as
end of the survey that asked them to type in short answers
they distribute limited monies. All of these efforts ultimately
to questions related to other important areas for oncology
can assist in meeting the mission of ONS: to promote excel-
nursing research and, in particular, to list three of the most
lence in oncology nursing and quality cancer care.
important research topics for oncology nursing to address in
the next five years. The enthusiastic feedback from the general
Author Contact: Ann M. Berger, PhD, RN, AOCN®, can be reached
membership who responded is perhaps an outward sign of a
at aberger@unmc.edu, with copy to editor at rose_mary@earthlink
thriving organization. Respondents were willing to take the
.net.
References
Cohen, M.Z., Harle, M., Woll, A.M., Despa, S., & Munsell, M.F. (2004).
agenda 2003­2005. Retrieved September 1, 2004 from http://www.ons.
Delphi survey of nursing research priorities. Oncology Nursing Forum,
org/research/information/documents/pdfs/2003Plan.pdf
31, 1011­1018.
Oncology Nursing Society. (2004, January). Summary report of 2003 ONS
Dillman, D. (2000). Mail and Internet surveys: The tailored design method
member surveys. Pittsburgh, PA: Author.
(2nd ed.). New York: Wiley.
Ropka, M.E., Guterbock, T.M., Krebs, L.U., Murphy-Ende, K., Stetz, K.M.,
Funkhouser, S.W., & Grant, M.M. (1989). 1988 ONS survey of research
Summers, B.L., et al. (2002). Year 2000 Oncology Nursing Society Re-
priorities. Oncology Nursing Forum, 16, 413­416.
search Priorities Survey. Oncology Nursing Forum, 29, 481­491.
Grant, M.M., & Stromborg, M. (1981). Promoting research collaboration: ONS
Schonlau, M., Fricker, R.D., & Elliott, M.N. (2001). Conducting research sur-
Research Committee survey. Oncology Nursing Forum, 8(2), 48­53.
veys via e-mail and the Web [Online book]. Rand Corporation. Retrieved
McGuire, D.B., Frank-Stromborg, M., & Varricchio, C. (1985). 1984 ONS
May 17, 2004, from http://www.rand.org/publications/MR/MR1480
Research Committee survey of membership's research interests and in-
Stetz, K.M., Haberman, M.R., Holcombe, J., & Jones, L.S. (1995). 1994
volvement. Oncology Nursing Forum, 12(2), 99­103.
Oncology Nursing Society Research Priorities Survey. Oncology Nursing
Mooney, K.H., Ferrell, B.R., Nail, L.M., Benedict, S.C., & Haberman, M.R.
Forum, 22, 785­789.
(1991). 1991 Oncology Nursing Society Research Priorities Survey. On-
Vehovar, V., Batagelj, Z., Manfreda, K.L., & Zaletel, M. (2002). Nonresponse
cology Nursing Forum, 18, 1381­1388.
in Web surveys. In R.M. Groves, D.A. Dillman, J.L. Eltinge, & R.J. Little
Oncology Nursing Society. (2003). Oncology Nursing Society research
(Eds.), Survey nonresponse (pp. 229­242). New York: Wiley.
ONCOLOGY NURSING FORUM ­ VOL 32, NO 2, 2005
290