Translating Resuscitation Guidelines into Practice: Health Care Provider Attitudes, Preferences and Beliefs Regarding Pediatric Fluid Resuscitation Performance

Introduction Children who require fluid resuscitation for the treatment of shock present to tertiary and non-tertiary medical settings. While timely fluid therapy improves survival odds, guidelines are poorly translated into clinical practice. The objective of this study was to characterize the attitudes, preferences and beliefs of health care providers working in acute care settings regarding pediatric fluid resuscitation performance. Methods A single-centre survey study was conducted at McMaster Children's Hospital from January to May, 2012. The sampling frame (n = 115) included nursing staff, physician staff and subspecialty trainees working in Pediatric Emergency Medicine (PEM) or Pediatric Critical Care Medicine (PCCM). A self-administered questionnaire was developed and assessed for face validity prior to distribution. Eligible participants were invited at 0, 2, and 4 weeks to complete a web-based version of the survey. A follow-up survey administration phase was conducted to improve the response rate. Results Response rate was 72.2% (83/115), with 83% (68/82) self-identifying as nursing staff and 61% (50/82) as PCCM providers. Resuscitation experience, frequency of shock management, and years in specialty, were similar between PCCM and PEM responders. Physicians and nurses had differing opinions regarding the most effective method to achieve rapid fluid resuscitation in young children presenting in shock (p<0.001). Disagreement also existed regarding the age and size of patients in whom rapid infuser devices, such as the Level-1 Rapid Infuser, should be used (p<0.001). Providers endorsed a number of potential concerns related to the use of rapid infuser devices in children, and only 14% of physicians and 55% of nursing staff felt that they had received adequate training in the use of such devices (p = 0.005). Conclusions There is a lack of consensus among health care providers regarding how pediatric fluid resuscitation guidelines should be operationalized, supporting a need for further work to define best practices.


Introduction
Shock is a life-threatening condition for infants and children that requires immediate resuscitative intervention. [1][2][3][4][5] While the actual incidence of pediatric shock is unknown, it is a worldwide problem, with causes including severe dehydration, bleeding, infection and allergy. With the exception of cardiogenic shock, circulatory instability generally results from conditions of impaired preload due to a state of relative or absolute intravascular hypovolemia. Acute intravascular volume loading increases end diastolic pressure, which in keeping with Frank-Starling principles improves stroke volume and cardiac output. [6] Current resuscitation guidelines require the timely performance of intravascular fluid administration as a critical initial pediatric shock management as this has been shown to improve survival odds. [7][8][9] Recent reports have led to questions about the role of aggressive fluid resuscitation in pediatric septic shock. [10,11] However it is imperative to note that these findings relate to children in the developing country setting where malnutrition, malaria, and anemia are common if not endemic. Such factors likely have an important influence on septic shock physiology including the response to fluid therapy. [12][13][14][15] As such, it is far from clear that these results should be extrapolated to children with septic shock in the developed country setting, or children with other forms of shock residing anywhere. Fluid administration therefore remains a 'cornerstone' of initial pediatric shock management.
Translating current fluid resuscitation guidelines into practice remains a challenge. Large volumes of fluid may be required for the effective resuscitation of septic shock, including volumes of up to 100 to 200 mL/kg or more in the first hour. [16,17] The American College of Critical Care Medicine (ACCM) guideline for the management of pediatric and neonatal septic shock calls for the administration of up to 60 mL/kg of isotonic fluid within the first 15 minutes of resuscitation. [2] While recommended benchmarks are often not achieved, [8,18,19] little is known about barriers to optimal pediatric fluid resuscitation performance.
It is important to consider that there are practical and technical challenges associated with simply getting the fluid into the patient when this is emergently required. While the impact of intravenous (IV) catheter [20][21][22][23] and tubing set [24,25] properties on fluid flow rates have been studied extensively, provider-level factors have not. To date, a single study has evaluated fluid resuscitation efficiency in pediatric patients in the non-operative setting. [26] A traditional infusion pump (Figure 1), gravity flow, manual syringe techniques (Figures 2 and 3), pressure bag support (Figure 4), or a rapid infuser device ( Figure 5) may be used to deliver fluid therapy. Traditional infusion pumps are capable of administering fluid at a maximum rate of 999 mL/hour, which is inadequate for the purposes of resuscitation. In contrast, rapid infuser devices such as the Level 1 (Level 1 H-1200 Fast Flow Fluid Warmer), trademark of Smiths Medical family of companies (Smiths Medical ASD, Inc., Rockland, MA, USA), [27] can administer fluid at high flow rates, however such devices are costly, require operator training, and may not be readily available.
It has been our observation in the non-operative clinical setting that health care providers use a variety of fluid administration techniques in the resuscitation of pediatric patients with decompensated shock -a recognized medical emergency. Given a lack of specific direction in current guidelines as to how fluid administration is best accomplished, we sought to evaluate the attitudes and beliefs of health care providers working in acute care areas regarding pediatric fluid resuscitation performance. We hypothesized that differing attitudes and beliefs would exist among providers suggesting a need for further research to define best practices.

Ethics Statement
The Faculty of Health Sciences/Hamilton Health Sciences Research Ethics Board approved this study and all procedures were conducted in accordance with the Tri-council Policy Statement: Ethical conduct for research involving humans. [28] A letter of information/consent was provided to all individuals invited to participate in the study. Written consent was not obtained from Phase 1 participants who completed the on-line version of the survey, as this was impracticable. Written consent was obtained and kept on file for Phase 2 study participants who were directly approached and completed the tablet-based version of the survey. Consent procedures were conducted according to the recommendations/requirements of our local Research Ethics Board.

Study Participants
This single center survey study was conducted at McMaster University Medical Centre (Hamilton, Canada) between January and May 2012. Our sampling frame included all staff nurses (full time, part time and occasional staff), staff physicians (full time, part time and occasional staff), and subspecialty physician trainees working in the Pediatric Emergency Department (PED) or Pediatric Critical Care Unit (PCCU) of McMaster Children's Hospital (n = 115). We chose to survey providers working in the PED and PCCU settings specifically because this group would be expected to be most likely involved in performing pediatric resuscitations. The sampling frame was developed based on information provided by key administrative staff with knowledge of and access to nursing and physician staff lists for these areas. Excluded from participation were medical students and resident trainees other than those pursuing subspecialty training in either Pediatric Critical Care or Pediatric Emergency Medicine. A good command of English was required to participate.

Survey Development
The survey was developed by the investigators to fulfill the study objectives. Questions elicited demographic information about participants and solicited their attitudes, experiences and beliefs regarding the performance of emergency fluid resuscitation for shock in children of different ages/sizes using different equipment and techniques. Questions also explored attitudes and beliefs concerning safety considerations. A draft version of the question- naire was assessed for face validity in 5 subjects, who provided verbal and written feedback to the investigators. Comments and suggestions were reviewed and incorporated into the final version of the questionnaire. [29] The finalized version of the survey permitted participants to skip over questions for which they did not wish to provide a response.

Survey Administration
The protocol involved administration of the survey in two distinct phases. In the first phase, eligible participants were invited via email to complete a web-based (Survey Monkey) version of the survey at 0, 2, and 4 weeks. Participants were instructed to complete the survey only once. A chance to win a $25 coffee card was advertised as an incentive. The protocol included provisions to proceed with a second phase of survey distribution if a response rate of at least 70% had not been achieved by 1 month following the final email inviting study participation.
In the second phase of survey administration, an email was sent to all individuals in the sampling frame alerting them that a Steps A through C are repeated as quickly as possible until the desired volume of fluid has been administered. This method of fluid administration is not sophisticated, but is commonly practiced in our experience. doi:10.1371/journal.pone.0058282.g002 research assistant would be visiting the PED and PCCU over a 2week period to offer a final opportunity to complete the survey. A research assistant visited the PED and PCCU daily and made contact with individuals who appeared available to be approached. Eligible and consenting participants who had not previously completed the survey were presented with an Apple iPad2 tablet device (32 GB with Wi-Fi, Apple model A1395), trademark of Apple Inc. (Apple Inc., Cupertino, CA, USA), to facilitate completion of the survey. The tablet-based version of the survey used during this phase was created using, iSURVEY (iSurveySoft, Wellington, NZ).

Study Outcomes, Data Management, and Statistical Methods
The primary outcome of this study was participant self-reported attitudes and beliefs regarding pediatric fluid resuscitation practices. As per our a priori plan, we also evaluated the potential impact of practice location (PED vs. PCCU) and profession (physician vs. nursing staff) on survey responses. The sample size was dictated by the number of eligible participants at our medical centre. Data from submitted surveys were temporarily stored on the Survey Monkey and iSURVEY websites. At study completion, data were downloaded onto the password-protected computer of the primary investigator (MP) and combined into a single data set for analysis. IBM SPSS statistics, Version 20 (IBM, Armonk, New York, USA) was used to perform all statistical analyses. Participant characteristics and survey data are summarized using mean (standard deviation) for continuous variables and count (percent) for categorical variables. Fisher's Exact Test (two-tailed) was used to assess for between group differences in proportions. In all cases significance was determined at the p,0.05 level.

Results
The survey response rate was 72% (83/115), with 83% (68/82) self-identifying as nursing professionals and 61% (50/82) as health care providers working in the PCCU setting. Seventy-six percent (63/80) of respondents reported that they worked full time and 56% (47/83) described themselves as experienced or very experienced in performing pediatric resuscitation. Additional details of participant characteristics are listed in Table 1.  Fifty-eight percent (46/80) of respondents were able to correctly identify that the ACCM pediatric and neonatal septic shock resuscitation guideline requires the administration of 20 mL/kg of isotonic fluid within 5 minutes. Study participants endorsed varying beliefs regarding how to optimally perform pediatric fluid resuscitation for children in shock (Tables 2 and 3). Opinions also differed regarding the factors most important in determining the mode of pediatric fluid resuscitation selected by providers (Table 4).
Less than half of those surveyed (47%; 39/82) believed that they had received adequate training/education regarding use of the Level 1 rapid infuser device in children. Differences of opinion also existed regarding the appropriate threshold patient size for use of this device, with physician respondents considering use in smaller patients more acceptable than nurses (p,0.001). Most respondents cited infrequent use of rapid infuser devices in their clinical practice (Table 5). Specific concerns cited by participants in relation to rapid infuser device use in children are listed in Tables 6  and 7.

Discussion
This study of PED and PCCU health care providers reveals differing attitudes and beliefs regarding optimal pediatric fluid resuscitation performance in the non-operative setting. Our findings are of potential importance in light of data indicating a frequent failure to meet recommended fluid resuscitation benchmarks for pediatric shock management in clinical practice.  [8,18,19] Lack of timely fluid therapy negatively impacts on survival odds [7][8][9] and increasing adherence to existing guidelines may help to improve outcomes. Understanding provider perspectives represents an important first step towards identifying and addressing knowledge gaps and barriers.

Fluid Resuscitation Preferences differ by patient age
While use of different fluid administration strategies in patients of differing ages/sizes may seem intuitive, respondents in our study still disagreed when presented with discrete patient age categories. In an in vivo study comparing the ability of gravity, push-pull, and pressure bag fluid administration techniques to meet the ACCM guideline goal of 20 mL/kg within 5 minutes, no technique succeeded in patients weighing 40 kg or more. [26] Gravity flow was inadequate regardless of patient weight, pressure bag was variably successful in patients under 40 kg, and the push-pull technique met the ACCM goal in all but one patient under 40 kg who also had a 22-gauge IV catheter in place. Consistent with these findings, respondents in our study cited manual fluid resuscitation methods as being preferred for patients up to 8 years of age. Our finding that factors independent of patient age/size appear to contribute to practice variation warrants further exploration. A lack of clear direction in current guidelines as to how fluid administration is best performed likely reflects the limited evidence available and this may well contribute to the variable preferences reported. Given that there appears to be room for improvement in fluid resuscitation performance in the clinical setting, further research to support development of evidence-based guidelines defining best practices seems sensible. Future guidelines should acknowledge patient size as important when choosing a fluid administration method.

Fluid Resuscitation Preferences differ by provider role
It is interesting to note the differences of opinion between physician and nursing staff regarding optimal fluid resuscitation performance. This finding suggests a possible disconnect between theory and practice that may apply to the performance of other resuscitative tasks as well. While physicians may have a good understanding of resuscitation requirements and endpoints, important knowledge concerning the practical aspects of implementation may be lacking. An illustrative example from our study relates to the perceived utility of rapid infuser devices. While physicians may have knowledge of the technical capabilities of these devices, nurses likely have a better understanding of the required setup time, personnel requirements, and operating complexities which may impact treatment initiation.
Providers are uncertain of the role of rapid infuser devices in pediatric fluid resuscitation Rapid infuser devices, such as the Level 1, are sophisticated medical devices that can be used to infuse isotonic fluids and/or blood products at high flow rates of up to 500 mL/min. This renders these devices extremely useful in the resuscitation of adult-  Table 4. Factors determining the fluid resuscitation method selected by health care providers when managing a pediatric patient in shock.  size trauma patients where the administration of litres of fluid may be required to rapidly restore circulatory volume. The role of rapid infuser devices in the resuscitation of pediatric patients is less clear, and the responses received from study participants reflect this uncertainty. Providers expressed differences of opinion regarding the threshold age/size of patients in whom use of rapid infuser devices should be considered, with physicians more open to using these devices in small patients. The limited weight-based data available from the Stoner study [26] would suggest that use of rapid infuser devices should be considered for patients weighing 40 kg or more, although other fluid administration techniques may be effective when more than one IV access site is available. At our pediatric tertiary care centre, Level 1 rapid infuser devices are located and ready for use in both the PED and the PCCU. This device is used periodically during both pediatric resuscitations and 'mock codes' at the discretion of the responsible physician.
In keeping with known complications that may occur with rapid infusion, [30][31][32] respondents indicated concern about the potential occurrence of adverse events in their patients. Participants also reported differing frequencies of use of rapid infuser devices in the clinical setting, with infrequent use commonly cited. Low frequency of use combined with provider concerns related to knowledge upkeep indicate a need for ongoing in-servicing of staff. Physicians may also benefit from education regarding device setup time, operating complexities, and personnel requirements.

Practical considerations may be most important in driving provider fluid resuscitation choices
More than half of the respondents in our study cited practical considerations as being most important in determining the mode of pediatric fluid resuscitation selected. In the operative environment, anesthetists can typically anticipate and prepare for the management of acute hypovolemia. In contrast, the non-operative environment is relatively less controlled and less predictable. Settings such as the PED and PCCU must therefore be broadly prepared to manage patients of all ages/sizes. Despite this, equipment may be subject to less frequent use, its location may change, and availability is dependent upon the adequate stocking of supplies. In other hospital areas, non-tertiary settings, and the prehospital setting, equipment availability may be even further limited. [33][34][35] The number of providers available to assist with resuscitation and their level of experience is also likely less consistent  in the non-operative setting. Current resuscitation guidelines may list a variety of options for performing fluid resuscitation recognizing that what works best in one setting may be impractical another.

Study Limitations and Strengths
As a single centre study, the main limitation to note relates to the sample size, which limits study power and may impact on the generalizability of our results. Our study also focuses on fluid administration practices used to address acute circulatory instability and does not delve into ongoing controversies regarding the most appropriate fluid type [36,37] or volume to administer beyond the initial resuscitative phase. [38,39] Study strengths include rigorous survey development and use of a staged protocol to maximize response rate. Our results suggest a need for further study of the practical aspects of fluid resuscitation performance to determine how health care providers can best achieve recommended benchmarks.

Conclusions
Differences of opinion regarding how best to perform pediatric fluid resuscitation exist among healthcare providers most likely to resuscitate children in the non-operative setting. Further study is warranted to define best practices and to improve guideline translation into the clinical setting. The role of rapid infuser devices in the resuscitation of pediatric shock requires further clarification.

Acknowledgments
Credit: iPad is a trademark of Apple Inc., registered in the U.S. and other countries.
Credit: Level 1 is a trademark of the Smiths Medical family of companies, registered in the U.S. and other countries.
Credit: iSURVEY is a copyrighted software program of iSurveySoft, Wellington, New Zealand.
Credit: We would like to thank Ms. Sara Urbanski for her assistance with data collection in this study.
Disclaimer: PLOS ONE is an independent publication and has not been authorized, sponsored, or otherwise approved by Apple Inc.