Conceptual framework. The Theory of Reasoned Action (TRA)21 and Theory of Planned Behavior (TPB)22 provided the theoretical foundation for this pilot study. Both theories focus on factors that impact the driving forces leading to the adoption of behaviors related to the use of the SCI-PUMT in clinical decision making. The TRA was developed by Fishbein,21 who differentiated between attitudes toward objects and attitudes toward behaviors directed toward those objects. Fishbein and Ajzen22 emphasized the importance of focusing on behavior (eg, using the SCI-PUMT) as the best attitude predictor.
In this pilot study, the perceived likelihood of adopting the behaviors of completing the SCI-PUMT and using this tool to make decisions regarding PrU treatment in veterans with SCI were assessed using statistical analysis. Behavioral intent was integral to performing these dual behaviors. Researchers presumed clinician intent to use the SCI-PUMT would be greater if study participants believed the SCI-PUMT was of value in assessing PrU healing and could assist in guiding PrU treatment. Thus, a positive attitude toward the behavior would translate into more positive use of the SCI-PUMT in clinical practice.23 The TPB augments the TRA by adding the dimension of perceived behavioral control (overall control of the behavior of using the SCI-PUMT in clinical practice).24 For example, clinicians may have been more likely to use the tool if they believed they had more autonomy in deciding this was the tool of choice for assessing PrU healing.
Setting. The study was conducted at the Michael Bilirakis SCI/D Center in Tampa, FL. The 100-bed center is comprised of 60 beds on 2 inpatient subacute care units, a 30-bed community living center, and a 10-bed unit for veterans who are ventilator-dependent. The Center also has outpatient and home care programs.
Approvals. This pilot study was approved by the University of Alabama’s Institutional Review Board (IRB) (IRB# 13-OR-177-ME) as a research study. It was approved by VHA Research and Development as a quality improvement project.
Data collection and analyses. Data collection forms were constructed for both phases using a TeleForm format (Cardiff TeleForm, Brookline, MA). TeleForm is a software application that automatically captures, classifies, and extracts data from documents by using recognition technologies. The purpose of using this software was to develop templates specific to this study that were printed, used for data collection, and subsequently scanned into an electronic file. Data were verified and then exported to an Excel file for data analysis. All quantitative analyses were performed using Statistical Analysis System software (SAS 9.3, SAS Institute Inc, Cary, NC).
Phase I: beliefs about using the SCI-PUMT.
Participants. To evaluate the intent to use the SCI-PUMT in clinical practice, a sample was selected of 5 physicians and 3 advanced registered nurse practitioners (ARNPs) (ie, “providers”) and 3 CWCNs (11 participants in total). This volunteer convenience sample represented the majority of providers and all of the CWCNs in the SCI/D Center. Providers and CWCNs were informed regarding the minimal risk and potential benefits for participating in the study via a letter approved by University of Alabama’s IRB. Completion of the questionnaire served as consent to participate in the study.
Processes/Procedures. Content and face validity of the Beliefs about Using the SCI-PUMT questionnaire was established by eliciting feedback from the 3 CWCNs whose primary responsibilities were to prevent, assess, and make recommendations for the treatment of veterans with PrU in the SCI Center. Validation focused on the design and content of the questionnaire. Modifications and/or item updates were incorporated into the final version before the initial phase of data collection (see Figure 2). Participants were asked to complete the questionnaire, which generated data regarding the intent to use the SCI-PUMT and attitudes, subjective norms, perceived behavioral control, and barriers related to these intentions.
This questionnaire was designed by the investigator to assess both direct and indirect measures of the TPB. The idea that behavioral intent is paramount in determining behaviors (ie, use of the SCI-PUMT) is integral to the tandem theories of TRA and TPB. It was assumed providers and CWCNs were more likely to use the SCI-PUMT if they believed outcomes associated with its use were positive.
The technical use of the SCI-PUMT to assess PrU healing in veterans with SCI was based on a semantic bipolar scale of 11 variables indicative of the intent to use the SCI-PUMT: 1 — extremely likely, 2 — slightly likely, 3 — neither likely nor unlikely, 4 — slightly unlikely, or 5 — extremely unlikely. The anticipated improvement in PrU assessment using the SCI-PUMT was based on a Likert scale of 3 variables that were indicative of the intent to use the SCI-PUMT: 1 — strongly agree, 2 — slightly agree, 3 — neither agree nor disagree, 4 — slightly disagree, or 5 — extremely disagree. Each end of the scale represented strong beliefs, and the midpoint represented no opinion (ie, “neither”). Negative aspects to these rating scales (eg, extremely unlikely, strongly disagree) were included because researchers believed attitudes also were shaped by negative attributes.
Statistical analysis. Due to the small sample and use of ordinal data, nonparametric statistics were used. Descriptive statistics were used for continuous variables (eg, mean, standard deviation, median) contained in the Beliefs about Using the SCI-PUMT questionnaire that was completed by providers and CWCNs. Descriptive statistics were calculated and analyzed for item, subscale, and total scale scores. Comments in the open-ended questionnaire were not thematically analyzed. No analysis was conducted to determine whether the individual providers or CWCNs who intended to use the SCI-PUMT actually did use it.
Phase II: SCI-PUMT and clinical decisions.
Sample. For Phase II of the study, the inclusion criteria were veterans enrolled in the SCI/D Center’s Registry who had a Stage II to Stage IV or unstageable PrU and had at least 2 SCI-PUMT scores on the same ulcer in April 2013. Exclusion criteria stipulated life expectancies of <6 months (ie, palliative care). The EHR was assessed to determine the participant population base in which the SCI-PUMT score appeared in the record. If the SCI-PUMT score was documented in the record, it was assumed the score was considered if the treatment did change. The sample was derived from 69 veterans who had PrUs included on their problem lists or had PrUs documented; 45 of these records were excluded because these did not have two or more SCI-PUMT scores during the study month. The final sample consisted of 24 veterans with 30 PrUs; 4 had multiple PrUs. These records were examined to determine the association between the magnitudes of change in SCI-PUMT total scores and changes recommended by the CWCNs for PrU treatment.25 Confidentiality was ensured by using identification numbers instead of names of providers, CWCNs, and veterans.
Processes/procedures. A SCI-PUMT Clinical Decisions Data Collection form was developed by the investigator to document SCI-PUMT scores and PrU treatments (see Figure 3). The CWCN could input information (document) on the SCI-PUMT in the EHR, but the providers wrote the orders regarding treatment decisions. Decisions were frequently based on the recommendations of the CWCN. This form listed 17 treatments with an option to add additional treatments not contained in the static list. Treatment was documented as a primary dressing (eg, hydrocolloid), debridement (eg, ultrasonic), or other intervention directly linked to PrU treatment. An inference was made that SCI-PUMT scores were considered in making clinical decisions about treatment based on whether the scores indicated that a plateau in healing had occurred (ie, score the same) or the PrU was getting worse (ie, score increased).
Data collection and statistic analysis. De-identified data were entered into a database, checked for errors, and stored on a secure server in the VHA computer system. Descriptive statistics were used for categorical variables (eg, frequencies, percentages) related to data extracted from the EHR regarding SCI-PUMT scores and PrU treatment. The independent variable was the SCI-PUMT score and the dependent variable was the change in PrU treatment. Scores from the SCI-PUMT and recommendations by the CWCN for treatment changes were entered into a 2 x 2 contingency table to calculate the chi-square.