Conducting a Large Survey of Critical Care Nurses in Canada:
Lessons Learnt
Elena Luk RN, BScN, CNCC(C); Louise Rose RN, BN, MN, PhD; Craig Dale RN, BScN, PhD(c), CNCC(C)
Introduction
• Survey research provides important information regarding:
• clinician opinion,
• adherence to evidence-based recommendations, and
• practice variation.
• Conduct of large surveys requires:
• considerable financial resources,
• time,
• energy,
• coordination, and
• ability to manage anticipated and unanticipated
logistical challenges.
Objective
• To inform nurse clinicians and researchers of our experiences in
the conduct of a large, Canadian survey of critical care nurses.
• We will illustrate administrative, financial, and logistical
challenges.
Survey Objective
• To survey critical care nurses across Canada to obtain a
description of their knowledge and practice of pain assessment
and management for critically ill adults.
Background
• Canada comprises 10 provinces and 3 territories.
• The majority of RNs (83%; 220,656/266,341) in Canada are
registered in the four provinces of Ontario, Quebec, British
Columbia, and Alberta.1
• Nursing associations in each province/territory maintain
regulation, with the exception of the Northwest Territories and
Nunavut that share an association.
• Each nursing association maintains databases that identify
nurses’ type of employment according to practice sector (e.g.
acute care hospital) and primary area of practice (e.g. critical
care).
Administrative Challenges
Determining the Sample Frame
• As only a small proportion of critical care nurses hold CACCN
membership (4.5% in 2008),2 we contacted the 12 nursing
associations to facilitate survey distribution through their
databases.
• One of our challenges was to identify the codes in the databases
that correctly identified nurses working in adult ICUs, as coding is
not the same across all nursing associations.
• We identified a total sample frame of 16,038 nurses.
• We used a stratified, disproportionate, random sampling
strategy to obtain representation of ICU nurses across all
provinces/territories; Total Sample = 3753 nurses.
• Inconsistencies in nursing associations’ databases challenged our
ability to achieve sample targets as a result of:
• Misclassification of nurses working in adult vs.
pediatric/neonatal ICUs.
• Registration lists containing names of nurses who had
moved, evidenced by surveys “returned to sender”, and
who identified themselves as no longer practicing in an ICU.
Corresponding with Nursing Associations
• Our first challenge was to determine the person(s) responsible
for granting and organizing research requests.
• We had to submit research applications to each association.
•
Financial Challenges
Logistical Challenges
The study was funded through a competitive peer reviewed
grant from the American Association of Critical Care Nurses
($15,000 USD) and the Nursing Research Fund of St. Michael’s
Hospital ($5,000 CAD).
Survey Tracking
Minimizing Survey Administration Costs
• Careful consideration was required to select location for printing
of unique identifiers to facilitated efficient matching of survey to
envelope eliminating the need to examine the master tracking
list with each envelope stuffing.
Postage
• To maximize response rates, we used three rounds of postal
contact with a thank-you/reminder letter and a replacement
survey sent at two-week intervals after initial survey.
• Postage for three separate mailings and reply-paid envelopes
comprised the largest expense.
• As domestic mail is charged according to weight, careful
planning was required to format survey in manner to stay
within lowest weight grade (<30g) to minimize costs.
Administering Surveys to Francophone Nurses
• Francophone nurses predominantly live in Quebec and New
Brunswick meaning these nurses needed to receive surveys in
their preferred language.
• New Brunswick’s nursing association could not identify
members’ preferred language, resulting in additional costs
due to the increased weight of mailing two surveys in
English and French.
• Translation services charge per word; Fortunately, our costs for
survey translation from English
French were minimized
as we had access to bilingual research assistants.
Survey Translation
• Some behavioural pain descriptors were not easily translatable
into French resulting in additional team discussion and expert
consultation to generate appropriate French descriptors.
•
Additional challenges were found when coding comments and
answers to open-ended questions to ensure understanding and
consistency.
Increasing Response Rates
• Despite three rounds of survey contact our response rate was
only 24.5% which compared unfavourably to the response rate
(57%) of pilot testing.3
• Due to funding limitations preventing further mail reminders,
two more reminders using e-mail and SurveyMonkeyTM were
sent to nurses from the seven nursing associations that were able
to release e-mail addresses, increasing the response rate by 1.7%.
Total Study Expenses
• Each association had different regulations regarding the
release of nurses’ contact information creating challenges in the
preparation, distribution, and tracking of survey materials.
• 5/12 (42%) nursing associations released mailing lists to us
• 6/12 (50%) did not release mailing lists but agreed to
distribute surveys for us
• Remaining nursing association in Prince Edward Island
(PEI) did not have administrative structure to assist with
survey distribution
• To overcome this challenge, we contacted nurse
managers of the two ICUs in PEI for assistance.
Conclusion
References
1. Canadian Institute for Health Information (2010) Regulated Nurses: Canadian Trends, 2005-2009, Ottawa, Ontario. Available at:
http://secure.cihi.ca/cihiweb/products/nursing_report_2005-2009_en.pdf (accessed February 24 2011).
2. Cybulski P & Mahon K (2009): Generations in the workforce at CACCN: results of CACCN 2008 membership survey. Dynamics 20, 7-10.
3. Rose L, Haslam L, Dale C, Knechtel L, Fraser M, Pinto R, McGillion M & Watt-Watson J (2011): Survey of assessment and management of
pain for critically ill adults. Intensive and Critical Care Nursing 27, 121-8.
RESEARCH POSTER PRESENTATION DESIGN © 2011
www.PosterPresentations.com
• Funding for critical care nursing research is limited therefore
careful planning is required to ensure surveys generate
important data with efficient use of resources.
• Potential survey participants should consider the significant
resources required to conduct surveys when deciding to respond
to survey invitations.