Executive Function Performance Test (Efpt) for Healthcare

¶ … Executive Function Performance Test (EFPT) for healthcare practitioners who deal with mild and moderate-severity stroke patients. Executive function is important to overall health because it allows us to regulate our actions, plan our behavior, and set goals. Stroke patients frequently suffer from executive-function related losses; this study proposes a measure that will help quantify the extent of executive function loss.

The design of this study was a basic experimental 3x1 design. The authors compared three groups on their performance of the EFPT. The three groups were: mild stroke, moderate stroke, and age-matched control (no stroke). All participants performed the EFPT and stroke group participants were tested 6 months after the stroke event to allow for an even amount of healing time.

The EFPT is noticeably different from other performance-based measures because unlike the Kitchen Task Assessment (its closest relative), it records cues from experimenters that support patient performance. Support cues may be offered as physical assistance or as verbal requests for assistance. Also, unlike other popular instruments that measure performance of day-to-day tasks, the EFPT formalizes this record of support cues as progressive -- just as in real life, caretakers or companions start by offering minimal support and progress to performing instrumental support (without which the task could not be done). This introduces a necessary element of realism to the measure. Especially with recovering stroke patients, the amount of help offered and accepted can be a critical distinction between different levels of post-stroke healing. The EFPT is better suited to accommodating people with motor impairment, whether or not they are stroke survivors, because those individuals routinely ask for and receive assistance with daily tasks.

The independent variables in this study were stroke condition (mild, moderate, or none), gender, age, education (years) and race (African-American or White). Two levels of dependent variable were described in this study: actual EFPT tasks (Cooking, Using Telephone, Medicine, and Paying Bills) and EF 'Components' (Initiation, Organization, Sequencing, Safety and Judgment, and Completion). Participants' scores were computed according to both sets of variables. The task variables were operationalized as the sum of scored/coded items during each task. The components variables were operationalized as the sum of items coded according to each component. Actions and events performed during the Using Telephone task, for example, could be coded as any one of the components, and each component was included in each of the EF tasks.

Reliability in the case of this instrument can be measured both through inter-coder reliability and internal consistency of the measure. The authors used Inter-Coder Coefficients as a measure of coder reliability and obtained high values for these ICCs on all EFPT tasks. The internal consistency was measured using Cronbach's alpha, for which a value of 0.94 was obtained, indicating good internal consistency. In other words, there was a high degree of correlation between participants' performance on each task.

Validity for this instrument has already been established for normal aging populations in other studies. For stroke populations, the construct validity of this instrument is not entirely consistent. For example, the instrument fails to effectively discriminate between mild and moderate stroke victims' ability in the Cooking task. In order to assess criterion validity, the authors compared the EFPT to a battery of neuropsychological tests. The EFPT's results corresponded highly with these neuropsychological tests, confirming the criterion validity of the instrument.

Table 1 presents the overall quantitative findings of the study in detail. Demographic variables are presented, followed by measured performance variables from the EFPT. In general, it appears that all of the EFPT variables produced significant differences between the Mild, Moderate, and Control conditions. However, the Moderate group contained only 14 individuals. Demographic proportions are about average for a study of this population, and no significant demographic differences were obtained between groups.