Children and adolescents with type 1 diabetes are often nonadherent with physician recommendations (Anderson, Svoren, & Laffel, 2007
; Greening, Stoppelbein, Konishi, Jordan & Moll, 2007
; Holmes et al., 2006
; Weissberg-Benchell et al., 1995
). The extent of nonadherence to the diabetes regimen is variable, ranging from 20% to 93% (Kovacs, Goldston, Obrosky, & Iyengar, 1992
; La Greca & Mackey, in press; Rapoff, 1999
; Wysocki, Buckloh, Lochrie, & Antal, 2005
). Each year, nonadherence with the treatment regimen leads to medical complications, hospitalizations and consequently higher healthcare costs; estimated costs of non-adherence in the United States are estimated to be as high as $300 billion per year (La Greca & Bearman, 2003
). Poor adherence adversely impacts health often resulting in increased morbidity and mortality, as well as excessive use of health care services (La Greca, 1990
; Lemanek, Kamps, & Chung, 2001
; Quittner, Espelage, Ievers-Landis, & Drotar, 2002
). Nonadherence is a significant concern not only for medical providers but also for mental health providers. For example, in an analysis of 91 psychological consultations received from a pediatric diabetes clinic, 62% were requested to address concerns related to poor adherence (Gelfand et al., 2004
Despite breakthroughs in diabetes treatment technology (e.g., insulin pumps, electronic blood-glucose meters) and empirical support for intensive insulin regimens, the benefits of these advances may be diminished when recommendations are not followed. For instance, nonadherence can negatively impact clinical decisions made by health care providers, e.g., prescribing insulin doses based on the available information may be problematic, if full insulin needs are not known due to undisclosed nonadherence. Furthermore, the demands of the diabetes regimen can be burdensome to children and their families—often psychosocial interventions aiming to improve diabetes self-care are unsuccessful (see Anderson et al., 2007
for a review). Thus, as children and adolescents receive increasingly intensive and complex medical regimens, it becomes even more critical to have psychometrically robust assessment methodology to monitor their regimen adherence.
There is no universal agreement on explicit standards for measuring adherence; measures of adherence range on a continuum from direct to indirect, each offering several advantages and limitations (for reviews, see: La Greca & Bearman, 2003
; La Greca & Mackey, in press; Quittner, Modi, Lemanek, Ievers-Landis, & Rapoff, 2008
; Wysocki, 2006
). To summarize, although direct observation of regimen behaviors can provide a highly specific, unconfounded method of assessment, the approach is impractical and cost prohibitive. Twenty-four hour recall interviews (Johnson, 1993
; Johnson, Silverstein, Rosenbloom, Carter, & Cunningham, 1986
) may offer the most comprehensive, multidimensional alternative. However, despite associations with hemoglobin A1c (HbA1c) (Freund, Johnson, Silverstein, & Thomas, 1991
; Johnson et al., 1992
; Reynolds, Johnson, & Silverstein, 1990
), the procedure can be difficult to implement in many healthcare settings due to resources and costs required to conduct multiple interviews and analyze complex data (McNabb, 1997
). A structured interview, such as Diabetes Self Management Profile (DSMP; Harris et al., 2000
) offers cost-effective and efficient alternatives for assessing adherence. However, although the DSMP is easier to administer than the 24 h recall, it still requires 20–40 min of the patient and parent's time. Additionally, given the semi-structured nature of the DSMP interview, interviewers must possess a comprehensive knowledge of the diabetes regimen to accurately and reliably administer the measure. For application within our prior research, clinicians required 5–15 h training prior to scoring reliably with senior clinicians.
Consequently, a more efficient tool for assessing diabetes adherence is necessary. Youth or parental-report of adherence, via paper-and-pencil administration, offers an alternative to the interview format (Anderson, Auslander, Jung, Miller, & Santiago, 1990
; Wysocki, 2006
). However, in a recent analysis of empirically based assessment for adherence, only 11 measures were found for all pediatric medical conditions (Quittner et al., 2008
). Alarmingly, only four measures met the criteria for “well-established”, which was defined as, “at least two research teams have published sufficient information evaluating the measure and establishing its strong psychometric properties” (Quittner et al., 2008
, p. 918).
One promising such adherence measure is the Self Care Inventory (SCI), a 14-item self- and parent- report measure of behaviors associated with the self-care of type 1 diabetes (La Greca, Swales, Klemp, & Madigan, 1988
; La Greca & Bearman, 2003
). The measure was developed by a pediatric psychologist with expertise in type 1 diabetes (Annette M. La Greca, PhD) for clinical and research purposes. Versions are available for adolescents and parents with appropriate wording for each (identical item content). Respondents report on their behavior over a 2-week interval using a 5-point Likert scale. Items on the SCI reflect the main components of the type 1 diabetes regimen, including: monitoring and recording glucose, administering and adjusting insulin, regulating meals and exercise, and keeping appointments. Although the SCI was developed prior to the Diabetes Control and Complications Trial; DCCT, 1993
), the same core regimen components remain at present (Silverstein et al., 2005
). However, psychometric properties of the SCI have yet to be examined in a sample of youth on more recent, intensive regimens.
Given the paucity of empirically based adherence assessment tools in pediatric populations, the aim of the current study was to evaluate the psychometric properties of the SCI with adolescents. Studies of the SCI with adults have been promising (Weinger, Butler, Welch, & La Greca, 2005
) however, detailed and readily available psychometric data for adolescents and parents are lacking. This study examines the reliability of the SCI, including internal consistency, test-retest data, and parent–child agreement. The validity of the SCI is also examined through comparisons with a previously established adherence measure (DSMP interview), frequency of blood-glucose monitoring, and metabolic control (HbA1c). Reliability and validity in a subset of the sample on intensive regimens will also be examined. Overall, this research sought to conduct an extensive analysis of the SCI's psychometric properties by an independent research group1
with expertise in pediatric diabetes with the aim of solidifying the SCI as an empirically “well-established” measure of adherence.