|Home | About | Journals | Submit | Contact Us | Français|
Patient problem solving and decision making are recognized as essential to effective self-management across multiple chronic diseases. However, a health-related problem-solving instrument that demonstrates sensitivity to disease control parameters in multiple diseases has not been established.
To determine, in two disease samples, internal consistency and associations with disease control of the Health Problem-Solving Scale (HPSS), a 50-item measure with 7 subscales assessing effective and ineffective problem-solving approaches, learning from past experiences, and motivation/orientation.
Outpatients from university-affiliated medical center HIV (N=111) and diabetes mellitus (DM, N=78) clinics.
HPSS, CD4, hemoglobin A1c (HbA1c), and number of hospitalizations in the previous year and Emergency Department (ED) visits in the previous 6 months.
Administration time for the HPSS ranged from 5 to 10 minutes. Cronbach’s alpha for the total HPSS was 0.86 and 0.89 for HIV and DM, respectively. Higher total scores (better problem solving) were associated with higher CD4 and fewer hospitalizations in HIV and lower HbA1c and fewer ED visits in DM. Health Problem-Solving Scale subscales representing negative problem-solving approaches were consistently associated with more hospitalizations (HIV, DM) and ED visits (DM).
The HPSS may identify problem-solving difficulties with disease self-management and assess effectiveness of interventions targeting patient decision making in self-care.
Facilitating effective patient self-management of chronic disease is a challenge for health care professionals and patients.1,2 Patient problem solving and decision making are deemed important for translating education and adherence counseling into active self-management in daily life.3–5 Studies have identified patient problem solving as a predictor of coping and health behaviors in multiple diseases,6,7 and a recent meta-analysis found social problem-solving therapy effective in relieving emotional and physical symptoms in samples treated for depression, substance abuse, and cancer.8
Limitations in literature warrant further investigation. First, problem-solving treatments in primary care settings generally have not measured patient problem solving to screen patients for problem-solving intervention, or to evaluate whether interventions were effective in improving problem-solving ability.9–12 Second, there are currently no measures of patient health-related problem solving for use across chronic diseases or for patients with multiple health conditions. Although generic measures (Problem-Solving Inventory13 and Social Problem-Solving Inventory14) are available, generic scales are generally less sensitive to disease-specific clinical outcomes than disease- or health-specific measures.15–17 Finally, although improved disease control is often a desired intervention outcome, studies of associations between patient problem solving and clinical measures of disease control are few.
The purpose of this cross-sectional study was twofold: (1) to determine whether a health-related problem-solving measure (Health Problem-Solving Scale [HPSS]) demonstrates adequate internal consistency in two independent adult, outpatient, chronic disease samples (HIV, diabetes mellitus [DM]); and (2) to determine associations of the HPSS with selected disease control parameters (CD4, hemoglobin A1c [HbA1c], Emergency Department [ED] visits, and hospitalizations).
The HPSS was developed from two previous studies,5,18 and item and response formats were modeled after a standardized generic measure, the Social Problem-Solving Inventory.14 Based on a conceptual model of problem solving in chronic disease self-management, the HPSS assesses self-report of effective and ineffective approaches to managing health-related problems.5 Items were generated from a focus group study with adult diabetes patients, in which 251 patient descriptions of how they handle problem situations with self-management were coded then reviewed and categorized by an expert panel.18 Responses that differentiated participants in good versus poor disease control were developed into questionnaire items and grouped into subscales reflecting the coded categories reviewed by the expert panel. For the HPSS, wording for items used “health” or “health condition(s)”to refer to medical conditions rather than disease-specific terms.
The resulting HPSS had 56 items, 6 of which were dropped from the scale because of confusing wording during pretesting on a small sample. The final HPSS was a 50-item instrument with seven subscales representing coded categories from the conceptual model: (1) effective problem solving (EPS, e.g., “Before I do something that will affect my health, I think about all the options”), (2) impulsive/careless problem solving (IPS, e.g., “Sometimes, I know I am doing the wrong thing for my health problem, but I just cannot stop myself”), (3) avoidant problem solving (APS, e.g., “I put off trying to deal with problems with my health condition as long as I can”), (4) positive transfer of past experience/learning (PTR, e.g., “When I find something that makes my health condition better, then I make sure I remember what it was and how I did it”), (5) negative transfer of past experience/learning (NTR, e.g., “I get stuck doing the same things I have always done, even if those things do not help”), (6) positive motivation/orientation (PMO, e.g., “When I think about the problems that can happen because of poor health, I want to do even more to take care of my health condition”), (7) negative motivation/orientation (NMO, e.g., “I feel there is nothing I can do about problems that come up with my health”).
Respondents used a 5-point Likert scale ranging from “not at all true of me” (0 points) to “extremely true of me” (4 points). Subscale scores were calculated by summing scores for each item in the respective subscale. Health Problem-Solving Scale total score was derived using the formula: , which sums the subscale averages, with reverse scoring of the negative subscales. Higher subscale scores indicate more of that problem-solving characteristic. Higher total HPSS scores indicate more effective health-related problem solving.
Study Participants Outpatients were recruited from two university-affiliated medical center clinics selected to represent different populations with regard to disease and patient sociodemographic characteristics. The HIV sample was recruited from an HIV clinic whose patients are predominantly male, African-American, with Medicaid (60%). The DM sample was recruited from a Diabetes Center, whose patients are predominantly female, white, with 55% private insurance/self-pay (19% Medicaid). Eligibility criteria were ≥18 years of age, currently a clinic patient, with diagnosis of HIV or diabetes, able to give informed consent, and able to complete study questionnaires.
Procedure In clinic waiting rooms during routine appointment hours, a research assistant (RA) handed patients flyers describing the study. Interested patients were instructed to approach the RA to complete an informed consent followed by study questionnaires. Disease control data were gathered from participants’ electronic medical records. Participants received a $10 reimbursement. The study was approved by the Institutional Review Board.
Measures of Disease Control, Emergency Department Visits, and Hospitalizations For HIV and DM, CD4 count and HbA1c results, respectively, were obtained on the day of the clinic visit or the most recent date within 1 month of the clinic visit. The number of ED visits in the past 6 months and hospitalizations in the past year were self-reported and confirmed by medical record review. If patients underreported utilization data relative to the medical record, then medical record data were used. If participants overreported relative to the medical record, then the larger number was used to capture utilization that may have occurred outside the medical center.
Statistical Analyses Participant characteristics were analyzed using descriptive statistics. Gender, age, race/ethnicity, and disease duration were not associated with HPSS total or subscale scores in either sample; therefore, those variables were not included as covariates in further analyses. Cronbach’s alpha was used to determine internal consistency (reliability) of the HPSS total scale and subscales.19 Pearson’s product–moment correlation coefficients were used to measure associations between HPSS and CD4 and HbA1c. Because of departures from normality assumptions, Spearman’s rank correlation coefficients were used to measure associations between HPSS and hospitalizations and ED visits. Statistical analyses were conducted using SAS (Cary, NC, USA) statistical software package (version 9.1).
Participants were 199 patients (90% response rate in HIV clinic, 86% in DM clinic). Incomplete HPSS protocols (n=10, HIV) were excluded from analyses. Characteristics of the final HIV (N=111) and DM (N=78) samples are in Table 1.
Administration time for the HPSS was generally 5 minutes, with a few patients requiring up to 10 minutes. Health Problem-Solving Scale total and subscale scores are presented in Table 1. Cronbach’s alpha for the HPSS total scale and subscales, in the HIV and DM samples, respectively, were are follows: HPSS total scale 0.86 and 0.89, EPS 0.82 and 0.82, IPS 0.85 and 0.88, APS 0.82 and 0.89, PTR 0.69 and 0.74, NTR 0.82 and 0.85, PMO 0.63 and 0.66, and NMO 0.74 and 0.83. These internal reliability coefficients were comparable in the two samples and within an acceptable psychometric range.19
Consistent with the theoretical model,5 HPSS subscales reflecting effective/positive problem-solving components were intercorrelated (HIV sample r=0.70 to .73, all P<0.001; DM sample r=0.79 to .81, all P<0.001), and subscales reflecting ineffective/negative components were intercorrelated (HIV sample r=0.73 to .84, all P<0.001; DM sample r=0.78 to .85, all P<0.001). Positive subscales were weakly correlated with their respective negative subscales.
Associations between HPSS and CD4, HbA1c, hospitalizations, and ED visits were all in hypothesized directions in both HIV and DM (Table 2) (Fig. 1). Health Problem-Solving Scale subscales showed differential patterns of association in the two samples, with stronger associations of negative/ineffective subscales in HIV and stronger associations of positive/effective subscales in DM. Overall, associations were weaker in HIV than in DM.
In HIV and DM samples representing differing sociodemographic characteristics, the HPSS demonstrated good internal consistency and hypothesized associations with disease control variables. Differential associations of subscales with CD4, HbA1c, hospitalizations, and ED visits suggest that problem-solving skill, learning, and motivation components, may differ in their contribution to outcomes.14,17 This finding reinforces the importance of addressing multiple problem-solving components, not just rational problem-solving skill, for efficacy of problem-solving interventions.8
The current study demonstrates that the HPSS may have utility as a measure of patient health-related problem solving in multiple diseases. Analysis of the HPSS in larger and additional chronic disease samples is needed and encouraged to further determine psychometric properties of the scale and to reduce number of items for greater administration efficiency. Clinical utility of the HPSS includes use as a screening instrument to identify patients who may be most in need of problem-solving interventions as part of patient education and disease self-management training. Moreover, future reports on the scale’s sensitivity to change are warranted to establish its utility as an outcome measure for evaluating effectiveness of interventions in modifying patients’ problem-solving abilities for better disease management and control.
The authors wish to thank Bill Ruby, MD, Heather Campbell, MPH, Christopher Saudek, MD, and Sherita Hill Golden, MD, MHS for providing access to clinic patients recruited for participation in this study. This research was supported by a supplement to NIH grant R01 DK48117, grant NIH K01 HL076644, and the Johns Hopkins Center for Mind Body Research (R21 NS048593).
Potential Financial Conflicts of Interest None disclosed.
Please read the statements below. They are thoughts and feelings people sometimes have about taking care of their health condition(s) and problems that can come up with their health. Problems include trouble managing symptoms (like pain, fatigue, shortness of breath, or thirst) and difficulties sticking to your treatments (like taking medication, following a healthy diet, or doing exercise/physical activity). Problems with your health condition(s) can also include things like dealing with emotions, social relationships, or problems taking care of your health while trying to go about your everyday life.
Write down the number for how much these statements are true of how you think or feel. Please use these options:
Scoring of HPSS subscales and total Scale:
Effective Problem Solving (EPS) = Item nos. 7, 9, 13, 20, 21, 24, 28, 44, and 49
Impulsive/Careless Problem Solving (IPS) = Item nos. 8, 12, 27, 30, 38, 39, 41, and 42
Avoidant Problem Solving (APS) = Item nos. 4, 25, 31, 33, 43, 46, and 48
Positive Transfer/Learning (PTR) = Item nos. 5, 16, 19, 45, and 50
Negative Transfer/Learning (NTR) = Item nos. 2, 3, 10, 14, 17, 23, 26, 29, 35, 37, and 40
Positive Motivation/Orientation (PMO) = Item nos. 1, 6, 18, 32, and 36
Negative Motivation/Orientation (NMO) = Item nos.11, 15, 22, 34, and 47