Use of modified Delphi introduces the risk of chronological bias during clinical research interventions

BACKGROUND

A study aimed at reducing the time spent on the phone obtaining insurance preauthorisation in a neurosurgical clinic was successfully completed. However, the researchers were unable to reject the null hypothesis because of a combination of chronological bias and the Hawthorne effect.

AIM

To increase nurse researchers' awareness of the potential to introduce a chronological bias as a confounder in clinical research and suggest potential alternative approaches to study design.

DISCUSSION

The researcher shared the study's purpose, design and outcome measure with the participants before collecting the baseline data. This enabled the participants to alter their practice before the intervention was implemented (a chronological bias) and change their behaviour surrounding the outcome (the Hawthorne effect).

CONCLUSION

The use of the Delphi method became a catalyst for change before the collection of baseline data, the combination of chronological bias and the Hawthorne effect affecting the study's results.

IMPLICATIONS FOR PRACTICE

Nurse researchers seeking to improve practice should collect baseline data before informing participants and consider the risks and benefits of blinding (concealment) surrounding the outcome.

Abstract P849: Effects of Standardized Stroke Protocols on Nursing Confidence and Knowledge in Low Volume Telestroke Emergency Departments

Introduction: Nurses and staff in Emergency Departments (ED) with low monthly case volumes have few opportunities to build confidence and solidify skills in acute stroke management. The Nursing-driven Acute Stroke Care (NAS-Care) study tested a workflow model with empowerment of ED bedside nurses, clear role assignments for team members, and standardized protocols including a predefined run sheet.

Methods: Seven Texas hospitals participated in this prospective, multisite, baseline-controlled study as part of the Lone Star Stroke Research Consortium. After three months of blinded baseline data collection, the following interventions were implemented: NIHSS certification, nursing education including mock stroke codes, and a standardized flowsheet for code organization and documentation (run sheet). Participating nurses were surveyed before and after implementation of this process.

Results: The study was completed at 6 hospitals, with 180 patients in the pre-intervention group and 267 in the post-intervention group. The study intervention was found to improve Door-to-ED provider and Door-to-CT metrics but not physician-dependent metrics, Door-to-Needle or Door-to-Provider times (Provencher et al, ISC 2020). Completed surveys were returned by 97 nurses (pre-intervention) and 57 nurses (post-intervention). There were significant increases in the following questions (10 point scale, p<.001): “I understand goals and processes of stroke code activation”, “stroke codes at my institution are completed efficiently”, and “stroke codes are nursing-driven.” In the post-intervention surveys, nurses reported that the NAS-Care protocol improved understanding (mean score 8.0 +/- 2.4 SD/10) and efficiency (8.2 +/- 2.4/10), and reported that they would recommend NAS-Care to be adopted at other institutions (8.8 +/- 2.1/10).

Conclusion: Standardized nurse-driven stroke protocols improved self-assessed knowledge and confidence for nurses in EDs utilizing telestroke, in addition to gains in staff-dependent stroke metrics.

Initial Reliability and Validity testing of the DESTINY Stroke and Large Vessel Screening Tool

Background: Screening is a valuable tool in emergency medicine for triaging, activation of protocols, and resource allocation. The Dallas Emergency Department Screening Tool to Identify Stroke (DESTINY) was developed to screen for all stroke subtypes, with improved sensitivity for posterior circulation stroke and large vessel occlusion (LVO). The main purpose of this study was to evaluate how reliably the tool could be taught and used by ED Nurses. Methods: This is Phase 1 of a multi-phase study to evaluate the DESTINY tool. A vascular neurologist retrospectively reviewed 409 stroke code activations between January and December of 2018. Thirty descriptive vignettes were composed: right anterior circulation stroke (n=5), left anterior circulation stroke (n=5), posterior circulation stroke (n=5), TIA and stroke mimic (n=6), non-LVO stroke (n=5), and hemorrhagic stroke (n= 4). Fifteen emergency medicine registered nurses used the DESTINY tool to screen the vignettes following the stroke education and DESTINY training module. Results: Individual Pearson Correlation Coefficients (PCC) were calculated for each nurse compared to the vascular neurologist (range: 0.43860 to 0.96966; mean: 0.8746 ± 0.1484). All individual PCC were statistically significant except for one [PCC of 0.43860 (p = 0.0153)]. Cronbach’s Alpha scores were very similar across multiple raters (standardized alpha for nurse raters: 0.9853). Omnibus test to compare DESTINY scores by stroke type demonstrated left anterior LVO and mimic/TIA stroke had few outlier observations, however, poorly differentiated hemorrhagic from ischemic stroke Conclusion: As demonstrated by high interrater reliability, the DESTINY tool can be effectively taught to ED nurses using clinical vignettes. The DESTINY tool is good at differentiating non-LVO, anterior LVO, mimics, and TIAs; while the ability to distinguish between anterior LVO/posterior LVO and ischemic/hemorrhagic stroke was sub-optimal. Utilizing the lessons learned in the early version of the DESTINY screening tool we expect that the high-reliability scores will continue to improve in future prospective studies.

Elevated blood pressure after craniotomy: A prospective observational study

PURPOSE

Close hemodynamic monitoring after craniotomy is routine given risk for post-operative hypertension, systemic and neurological complications. Patient and peri-operative variables associated with increased risk of post-craniotomy hypertension and complications are not well understood. Our analysis aims to estimate the incidence and prevalence of post-craniotomy hypertension, its time course, contributing factors, and post-craniotomy complications.

MATERIAL AND METHODS

This is a prospective study of patients admitted to the Neurosurgical Intensive Care Unit after an elective craniotomy. Variables associated with pre-surgical risk, demographics, and post-operative care were analyzed.

RESULTS

A total of 282 patients were included in the final analysis, 44% had pre-existing hypertension. Post-craniotomy hypertension was seen in 21%, with a higher incidence in patients with pre-existing hypertension (p < .001), smaller craniotomies (p = .0035), and increased use of analgesic medications (p < .001). History of hypertension was the only independent risk factor for post-craniotomy hypertension in a multivariate regression model. Patients who developed post-craniotomy hypertension, showed a significant increase in length of stay, number and duration of antihypertensive treatment. However, post-craniotomy hypertension was not associated with a higher incidence of other post-operative complications.

CONCLUSIONS

Development of hypertension after craniotomy is multi-factorial. In this prospective study, a prior history of hypertension was the only associated independent risk factor.

Do Cognitive Therapy Skills Neutralize Lifetime Stress to Improve Treatment Outcomes in Recurrent Depression?

Cognitive therapy (CT) is an efficacious treatment for major depressive disorder (MDD), but not all patients respond. Past research suggests that stressful life events (SLE; e.g., childhood maltreatment, emotional and physical abuse, relationship discord, physical illness) sometimes reduce the efficacy of depression treatment, whereas greater acquisition and use of CT skills may improve patient outcomes. In a sample of 276 outpatient participants with recurrent MDD, we tested the hypothesis that patients with more SLE benefit more from CT skills in attaining response and remaining free of relapse/recurrence. Patients with more pretreatment SLE did not develop weaker CT skills, on average, but were significantly less likely to respond to CT. However, SLE predicted non-response only for patients with relatively weak skills, and not for those with stronger CT skills. Similarly, among acute-phase responders, SLE increased risk for MDD relapse/recurrence among patients with weaker CT skills. Thus, the combination of more SLE and weaker CT skills forecasted negative outcomes. These novel findings are discussed in the context of improving CT for depression among patients with greater lifetime history of SLE and require replication before clinical application.

Abstract TMP105: Location Does Matter: Comparing the Location of Blood Pressure Management in an Intensive Care Unit Setting

Background and Purpose: Blood pressure (BP) management is a vital part of acute stroke care. Both over- and under-correction of BP are associated with increased morbidity and mortality. Non-invasive BP (NIBP) and intra-arterial BP (ABP) measurements are commonly used, yet both methods yield inconsistent results. Measurement error can hinder optimal clinical management in neurological conditions where BP directly affects cerebral perfusion pressures. Discrepancy in clinical decision-making associated with BP measurement is rarely reported. This prospective, non-randomized, cross-sectional study aims to address the gap by correlating simultaneous within-subject BP readings from multiple sites.

Methods: NIBP was simultaneously measured from 4 sites (both arms, both wrists) and ABP (when available) in 80 intensive care unit subjects. Correlation matrices and repeated measures ANOVA were used to explore for differences in BP by measurement site.

Results: Of 80 subjects, 41 were male, mean age = 52.8, and mean BMI = 30. Pearson Correlation Coefficients for SBP ranged from 0.67 to 0.83; DBP from 0.77 to 0.84; and MAP from 0.76 to 0.88. (Figure 1). SBP differences ranged from 0 to 57 mmHg and MAP differences ranged from 0 to 36 mmHg. One-way repeated measures ANOVA revealed significantly different values for SBP (p=0.0319); DBP (p=0.0002); and MAP (p=0.0001).

Conclusions: BP values vary significantly when measured simultaneously in different sites. Unpredictable inter-site comparisons of BP warrant significant research in larger prospective trials. Nurses should consider standardizing a measurement site for consistent BP readings and better clinical decision-making.

Patient perceptions and knowledge of Parkinson's disease and treatment (KnowPD)

Background

As the 2nd most common neurodegenerative disorder, Parkinson's disease (PD) affects over 1 million Americans. Treatment is complex and may include pharmacotherapy, rehabilitative measures, and surgical intervention. A comprehensive understanding of the patient's perceptions about PD is a vital step towards improving health literacy and clinical outcomes.

Methods

KnowPD is a web-based survey with Likert responses for a cross-sectional, nonrandomized study to assess patients' knowledge of PD symptoms, medications, deep brain stimulation (DBS), rehabilitation, and other factors relevant to disease management.

Results

Of the 98 subjects surveyed, 90% agreed they had sufficient knowledge of PD. However, in this cohort, 82% incorrectly believed levodopa stops working as the disease progresses, 77% erroneously thought DBS improves balance and reduces falls, and, <50% could answer specific questions regarding the dosing of levodopa despite over 75% reporting managing their own medications. A majority of patients (84%) believed it was possible to live well with PD, correlating with their self-reported knowledge of the disease. Finally, patients selected electronic video (13.7%) and reading (20.0%) material, yearly symposia (20.0%), and lunch lectures (28.4%) as their preferred method of information delivery.

Conclusion

Misconceptions are prevalent among PD patients, and these appear to be unrelated to gender, provider type, or education level. Identification and characterization of this knowledge gap is vital towards allocating patient education resources, and the findings described herein will form the basis for effective educational interventions.

Psychiatric Nurses Providing Post-Stroke Depression Education to Care Partners of Stroke: A Pilot Study

Background: Stroke survivors are a higher risk for post-stroke depression (PSD) after they return home from the hospital. Given that many stroke patients have friends or family members (care partners) present after hospital discharge, this pilot study examined the utility of psychiatric nurses educating family members on how to recognize PSD. Methods: This pilot study identified stroke patients and their care partner while they were in the hospital, to provide education on depressive symptoms, community resources, safety resources (e.g., suicide hotline), how to address symptoms of depression with their care partner. A descriptive observational design was used wherein a psychiatric delivered PSD education to the patient’s care partner (instead of the patient). Knowledge of depression was assessed immediately prior to-, immediately after-, and at a 2-week follow-up after-, the education intervention. Results: This pilot study was not powered to detect a statistically significant difference. However, among the 20 subjects (mean age 61 years), fewer scores fell in the lower range (less knowledge of depression) immediately after the education and at the 2-week follow-up phone call. Conclusion: This study demonstrates the potential of multi-disciplinary nursing consultation to improve outcomes.

Music as Medicine: The Therapeutic Potential of Music for Acute Stroke Patients

Nurses caring for patients with acute stroke are likely to administer both music and medication with therapeutic intent. The administration of medication is based on accumulated scientific evidence and tailored to the needs of each patient. However, the therapeutic use of music is generally based on good intentions and anecdotal evidence. This review summarizes and examines the current literature regarding the effectiveness of music in the treatment of critically ill patients and the use of music in neurologically injured patients. The rationale for hypothesis-driven research to explore therapeutic music intervention in acute stroke is compelling.

Pilot Study of Educational Gaming to Improve Adherence to an End-Tidal Carbon Dioxide Monitoring Protocol

BACKGROUND

End-tidal carbon dioxide (ETCO2) monitoring is an important part of patient care. Understanding and interpreting ETCO2 wavelengths can be a challenge. This pilot study explored the efficacy of a novel approach to educating clinicians on ETCO2 monitoring via game theory.

METHOD

A video game application for ETCO2 monitoring was developed. Clinicians were encouraged to play the game over a 3-month period. Compliance with the ETCO2 protocol was compared in a random selection of patients admitted before, during, and after the intervention.

RESULTS

Thirty-eight clinicians completed the preand posttest, with a significant difference in test scores (p = .03). The intervention was associated with higher adherence to the ETCO2 protocol before and after the intervention (p < .05).

CONCLUSION

The availability of new technologies has created opportunities to develop new approaches to educate clinicians. This study showed that the use of a game improved adherence to the ETCO2 protocol. J Contin Educ Nurs. 2018;49(2):79-83.

Use of Statins and Outcomes in Intracerebral Hemorrhage Patients

BACKGROUND AND PURPOSE

Statin use may be associated with improved outcome in intracerebral hemorrhage patients. However, the topic remains controversial. Our analysis examined the effect of prior, continued, or new statin use on intracerebral hemorrhage outcomes using the ERICH (Ethnic/Racial Variations of Intracerebral Hemorrhage) data set.

METHODS

We analyzed ERICH (a multicenter study designed to examine ethnic variations in the risk, presentation, and outcomes of intracerebral hemorrhage) to explore the association of statin use and hematoma growth, mortality, and 3-month disability. We computed subset analyses with respect to 3 statin categories (prior, continued, or new use).

RESULTS

Two thousand four hundred and fifty-seven enrolled cases (mean age, 62 years; 42% females) had complete data on mortality and 3-month disability (modified Rankin Scale). Among those, 1093 cases were on statins (prior, n=268; continued, n=423; new, n=402). Overall, statin use was associated with reduced mortality and disability without any effect on hematoma growth. This association was primarily driven by continued/new statin use. A multivariate analysis adjusted for age and major predictors for poor outcome showed that continued/new statins users had good outcomes compared with prior users. However, statins may have been continued/started more frequently among less severe patients. When a propensity score was developed based on factors that could influence a physician's decision in prescribing statins and used as a covariate, continued/new statin use was no longer a significant predictor of good outcome.

CONCLUSIONS

Although statin use, especially continued/new use, was associated with improved intracerebral hemorrhage outcomes, this effect may merely reflect the physician's view of a patient's prognosis rather than a predictor of survival.

Fertility preservation for cancer patients: Role of nursing attitudes

110

Background: Nurses play a pivotal role in ensuring adherence to quality guidelines. The 2006 American Society of Clinical Oncology - Quality Oncology Practice Initiative guidelines state that infertility risks should be discussed prior to initiation of fertility-reducing therapies for patients in their reproductive years. Despite this guideline on fertility preservation (FP), national compliance is only 6% (Tallent, 2013). Research is lacking in the understanding of nurses’ attitudes and behaviors towards FP and no tools exist from which to examine this problem. The purpose of this study is to gain increased understanding of potential barriers for nurses to recommend FP as an option for newly diagnosed patients. Methods: This instrument development study was developed by content experts and field tested with staff nurses. Next, an online survey method was used to examine factors associated with nursing recommendation for FP of newly diagnosed cancer patients. An anonymous online survey was sent to oncology nurses at UT Southwestern and members of the Oncology Nursing Society (Dallas Fort Worth Chapter). The endpoint of the study is a sound survey to assess nursing behaviors toward FP in diagnosed cancer patients. Standardized descriptive measures of central tendency were used to describe baseline characteristics for the sample. Factor analysis was completed using SAS v9.3 with orthogonal rotation. Scree plot was examined for factors and loading > 0.30 were retained. Results: Analysis of the first 67 returned surveys (Phase 1) supports a 15-item (5-factor) survey tool. Preliminary results suggested the need for education, referral availability, and system refinement. Phase 2 surveys of additional 200 was initiated in August 2015, with an expected close of enrollment date of December 2015. Conclusions: Cancer patients of childbearing age worry how treatment will affect their ability to have a child. This survey may be beneficial in identifying nursing attitudes regarding FP. All members of the care team should be sensitive to this concern and should recommend appropriate action. This study will identify nursing barriers to patient education, and will support system-wide implementation of a comprehensive FP program.

Abstract T P255: QCI-NASCAR - Quality Care Improvement with Nursing-driven Acute Stroke CARe

Objective: To observe the impact on stroke code time metrics after applying a “pit stop” model of bedside nursing for telestroke encounters.

Background: Despite the recent push for target treatment times in acute stroke codes, no guidelines exist for optimizing practices specific to stroke care via telemedicine. Effective telestroke is dependent on efficient data gathering by remote staff, and lengthy metrics for real-world telestroke often preclude timely tPA treatment. By co-opting “pit stops” as inspiration, an optimized nursing workflow for telestroke can be created on the following principles: Identification of Shared Goals; Organized Urgency with the Removal of Gatekeepers; Multi-personnel, Non-Sequential Processes; Focus on Defined Staged Roles; and Empowered Engagement/Responsibility.

Methods: The QCI-NASCAR protocol was implemented in Oct 2013, and data was collected prospectively on consecutive stroke code activations through Apr 2014 at St. Paul University Hospital (Dallas, TX), a telestroke spoke site. The nurse-driven protocol was reinforced by a paper checklist (i.e. “Driver Sheet”), which doubled as a data collection form. Timestamps were recorded in real time for: door time, MD at bedside, CT arrival, needle time, and/or code cancellation. The primary outcome was Door-to-CT (D2CT) times to reflect the portion of the stroke code most impacted by the nursing protocol.

Results: Mean D2CT times were: all cases (n=152, 33.2 min), intervention-eligible cases (n=71, 27.0 min), and thrombolytic-eligible cases (n=57, 22.2 min). A trend for lower D2CT times and standard deviations was noted in comparing the first half of the data (n=76, 38.04 ± 58.1 min) to the second (n=77, 27.8 ± 19.1 min; p<0.05). A similar pattern was noted in the subset of intervention-eligible cases: first half (n=36, 29.4 ± 37.4 min) vs. second half (n=35, 24.3 ± 18.6 min; p<0.05). IV tPA was administered 3 times, including an institutional best door-to-needle time of 32.0 min.

Conclusion: QCI-NASCAR demonstrates the feasibility of implementing a nursing-driven protocol for telestroke encounters. A larger, multi-institutional trial will demonstrate if such a protocol can significantly and reproducibly lower stroke code metrics to national guideline parameters.

Abstract W P255: Prior Use of Statins May Reduce Hematoma Growth in Intracerebral Hemorrhage Patients

Introduction: Higher incidence and growth of intracerebral hemorrhage (ICH) has been reported with primary hypocholesterolemia. Statins have been associated with improved outcome in ICH patients, although they lower low density lipoprotein (LDL) cholesterol. Our study explores whether lower admission LDL secondary to statin use is associated with admission ICH hematoma volumes (HV) or hematoma growth (HG).

METHODS: Data from the Ethnic/Racial Variations of Intracerebral Hemorrhage (ERICH) study, a multicenter case-control study designed to examine ethnic variations in the risk, presentation, and outcomes of ICH, were analyzed to explore the association of statin use and LDL with HV and HG. The Kruskal-Wallis test was used to examine differences in initial volume between groups, and the Wilcoxon Signed-Rank test was used to examine differences in HV between patients’ first two CT scans.

RESULTS: Among cases enrolled prior to January 2013, 544 (mean±SD age=60.9±14.1, 60.3% male) had both admission LDL data and CT scan within 24 hours of onset. Four groups were defined: 1) statin use and LDL<70, n=42 (7%); 2) statin use and LDL≥70, n=79 (14%); 3) no statin use and LDL<70, n=58 (10%); and 4) no statin use and LDL≥70, n=365 (67%). There was no difference in admission HV among groups (p=.69). Follow up (f/u) imaging was available for 350 (64.3%) cases: n for the 4 groups was 26 (7%), 47 (13%), 35 (10%), and 242 (69%). Compared with baseline CT, the non-statin users had significantly higher f/u HV (p<.0001), whereas f/u HV was unchanged for statin users (p=.529). Group analysis showed decreased HV in group 1 and increased HV in group 4 (Figure 1).

CONCLUSION: There was no association between LDL<70 with or without statin use and admission HV. When compared with non-statin users, statin users, especially with LDL<70, had reduced HG on f/u scans. Further studies are needed to explore these findings.