Michael A. Kallen

Having a fit: impact of number of items and distribution of data on traditional criteria for assessing IRT's unidimensionality assumption.

PurposeConfirmatory factor analysis fit criteria typically are used to evaluate the unidimensionality of item banks. This study explored the degree to which the values of these statistics are affected by two characteristics of item banks developed to measure health outcomes: large numbers of items and nonnormal data.MethodsAnalyses were conducted on simulated and observed data. Observed data were responses to the Patient-Reported Outcome Measurement Information System (PROMIS) Pain Impact Item Bank. Simulated data fit the graded response model and conformed to a normal distribution or mirrored the distribution of the observed data. Confirmatory factor analyses (CFA), parallel analysis, and bifactor analysis were conducted.ResultsCFA fit values were found to be sensitive to data distribution and number of items. In some instances impact of distribution and item number was quite large.ConclusionsWe concluded that using traditional cutoffs and standards for CFA fit statistics is not recommended for establishing unidimensionality of item banks. An investigative approach is favored over reliance on published criteria. We found bifactor analysis to be appealing in this regard because it allows evaluation of the relative impact of secondary dimensions. In addition to these methodological conclusions, we judged the items of the PROMIS Pain Impact bank to be sufficiently unidimensional for item response theory (IRT) modeling.

Low Rates of Hepatitis B Virus Screening at the Onset of Chemotherapy

PURPOSE Patients with hepatitis B virus (HBV) infection are at risk for reactivation after chemotherapy. Effective prophylaxis is available but depends on detection of prior infection. Previous studies have shown low screening rates, but no large-scale US studies have been conducted. We sought to determine predictors of screening and positive HBV test results in patients receiving chemotherapy. METHODS We conducted a retrospective cohort study of patients with newly diagnosed cancer who received chemotherapy between January 2004 and September 2007 at a comprehensive cancer center. We determined rates and predictors of screening for HBV infection with HB surface antigen (HBsAg) and antibody to hepatitis B core antigen (anti-HBc) tests as well as the prevalence and predictors of positive results. We explored rates of acutely elevated liver function tests and liver decompensation after chemotherapy. RESULTS Of 10,729 new patients who received chemotherapy, 1,787 (16.7%) underwent HBsAg or anti-HBc screening. Less than 20% of patients with HBV risk factors were screened, even though their odds of HBV infection were increased four-fold compared with those without risk factors. The prevalence of chronic HBV infection was 1.5%. whereas 7.4% had positive anti-HBc only. The strongest predictors of HBV screening were having a history of HBV infection, hematologic malignancy, and rituximab treatment (P < .001). Asian ethnicity was not a significant predictor of screening, despite being a strong and highly significant predictor of positive test results (P < .001). CONCLUSION HBV screening among patients with cancer is low, especially among those known to be at high risk for HBV infection. Future research directed toward identifying best screening methods and HBV risk tools will be necessary to reduce the risk of reactivation of HBV infection after chemotherapy.

Improving Wait Time for Chemotherapy in an Outpatient Clinic at a Comprehensive Cancer Center

PURPOSE We conducted our study at the Ambulatory Treatment Center (ATC) of the MD Anderson Cancer Center, a network of six outpatient treatment units for patients receiving infusion therapies. Excessive patient wait time for chemotherapy was a primary source of ATC patient dissatisfaction. ATC employees expressed frustration, because often, patients arrived physically on time but were not treatment ready. Additionally, ATC staff emphasized challenges associated with obtaining finalized treatment orders for prescheduled appointments (ie, placeholder appointments without associated physician treatment orders). We aimed to decrease mean patient wait time from check-in to treatment in one ATC unit by 25%. METHODS We studied appointment cycle time in the ATC Green Unit, stratifying appointments by type (ie, prescheduled [no finalized treatment orders] and scheduled [finalized treatment orders]). We obtained mean wait times at baseline (control) and again after our intervention period. We conducted interviews and observations in ATC Green, from which we developed a three-part plan to reduce wait time: increase process efficiency within ATC Green, enhance communications with MD Anderson clinics and centers, and incorporate information technology applications. RESULTS After our intervention, we observed a 15% decrease in wait time for patients with prescheduled appointments and a 29% decrease for those with scheduled appointments. Overall, there was a 26.8% reduction in mean patient wait time relative to baseline (control). CONCLUSION We observed a significantly decreased mean patient wait time after implementing our intervention. This decrease may improve patient satisfaction, relieve employee frustration with appointment scheduling, and create opportunities for increasing institutional revenue.

Impact of antiretroviral dosing frequency and pill burden on adherence among newly diagnosed, antiretroviral-naive HIV patients.

There are few data on the impact of antiretroviral therapy (ART) regimen factors on adherence in ART-naïve HIV patients on contemporary once- or twice-daily regimens. Ninety-nine newly diagnosed patients in a prospective observational cohort study completed a visual analogue scale to assess their ART adherence. Adherence by type of ART and dosing frequency were compared by Brown–Mood median tests. Participants taking once-daily regimens had higher adherence (n = 70, 99.5%) compared with participants taking twice-daily regimens (n = 29, 94%; P = 0.01). Adherence of participants taking the fixed dose combination efavirenz–emtricitabine–tenofovir (n = 34, 100%) compared with those taking once-daily regimens of two or more pills was no different (n = 36, 99.3%; P = 0.34). Among a cohort of newly diagnosed ART-naïve patients, once-daily dosing of ART resulted in higher adherence than twice-daily dosing. Pill burden among once-daily regimens did not predict adherence, suggesting that factors other than pill burden should drive regimen selection.

Validity of self-report measures in assessing antiretroviral adherence of newly diagnosed, HAART-Naïve, HIV patients

Abstract Purpose: To compare the performance of self-report instruments assessing adherence to antiretroviral therapy (ART) in patients starting ART for the first time and in a predominately Hispanic population.Methods: Of 184 patients in a prospective observational cohort study of newly diagnosed, minority patients of low socioeconomic status, 54 were given Medication Event Monitoring System (MEMS) caps for their boosted protease inhibitor (PI) or non-nucleoside reverse transcriptase inhibitor (NNRTI). They completed a 4’week recall visual analogue scale (VAS), the Adult AIDS Clinical Trial Group (AACTG) 4’day recall instrument, and a 1’month recall qualitative single-item measure every 3 months for up to 18 months in English or Spanish. Electronic pharmacy records recorded refill dates. Spearman correlation coefficients were calculated to compare self-report measures with MEMS data and pharmacy data.Results: Of 46 patients with MEMS data, mean adherence was 84.7% (SD 35.6) by MEMS, 84.5% (SD 15.1) by pharmacy, 95.4% (SD 11.9) by VAS, 95.8% (SD 17.2) by AACTG, and 87.6% (SD 28.2) by qualitative single item. The correlation coefficient (CC) of VAS with MEMS was 0.37 (P < .01), and with pharmacy it was 0.34 (P < .01). The CC of the AACTG with MEMS was 0.32 (P < .01), and with pharmacy it was 0.28 (P < .01). The qualitative single item had a CC with MEMS of 0.24 (P < .01) and with pharmacy of 0.32 (P < .01). Spanish-speaking patients’ VAS adherence had a CC of 0.40 (P < .01) with MEMS.Conclusions: The VAS, AACTG, and qualitative single-item measures correlated significantly with MEMS and pharmacy data. Our data support self-administration of the VAS, even in Span-ish speakers.

Quality appraisal of clinical practice guidelines and consensus statements on the use of biologic agents in rheumatoid arthritis: A systematic review

OBJECTIVE To evaluate the quality of clinical practice guidelines (CPGs) and consensus statements (CS) for the treatment of rheumatoid arthritis with tumor necrosis factor alpha (TNFalpha) antagonists. METHODS We searched for CPGs and CS on the use of infliximab, etanercept, and/or adalimumab for the treatment of rheumatoid arthritis, published through October 10, 2006. Sources included electronic databases (Medline, EMBase, BIOSIS, etc.), guideline registries, and pertinent Web sites. Review of 4,915 citations revealed 16 CPGs and 20 CS. Two independent reviewers evaluated development methods of selected studies using the 23-item Appraisal of Guidelines for Research and Evaluation (AGREE) instrument and compared recommendations between guidelines. RESULTS Of the 16 guidelines, only 5 (31%) were based on a systematic review of relevant research evidence. Only 4 (25%) of the guidelines fulfilled > or = 60% of the AGREE criteria. AGREE scores were lower for guidelines from rheumatology societies than government agencies when reporting scope and purposes (P = 0.03), stakeholder involvement (P = 0.03), and clarity and presentation (P = 0.01). Guidelines scored higher than CS in most domains. Overall, guideline recommendations were consistent with respect to the use of biologic agents after failure of disease-modifying antirheumatic drugs, but differed or did not provide specific guidance on tests for screening. CONCLUSION Guidelines for introducing TNFalpha antagonists in rheumatoid arthritis often fail to meet expected methodologic criteria and therefore vary significantly in quality and with respect to some recommendations for patient assessment and management.

Improving benchmarking by using an explicit framework for the development of composite indicators: an example using pediatric quality of care

BackgroundThe measurement of healthcare provider performance is becoming more widespread. Physicians have been guarded about performance measurement, in part because the methodology for comparative measurement of care quality is underdeveloped. Comprehensive quality improvement will require comprehensive measurement, implying the aggregation of multiple quality metrics into composite indicators.ObjectiveTo present a conceptual framework to develop comprehensive, robust, and transparent composite indicators of pediatric care quality, and to highlight aspects specific to quality measurement in children.MethodsWe reviewed the scientific literature on composite indicator development, health systems, and quality measurement in the pediatric healthcare setting. Frameworks were selected for explicitness and applicability to a hospital-based measurement system.ResultsWe synthesized various frameworks into a comprehensive model for the development of composite indicators of quality of care. Among its key premises, the model proposes identifying structural, process, and outcome metrics for each of the Institute of Medicines six domains of quality (safety, effectiveness, efficiency, patient-centeredness, timeliness, and equity) and presents a step-by-step framework for embedding the quality of care measurement model into composite indicator development.ConclusionsThe framework presented offers researchers an explicit path to composite indicator development. Without a scientifically robust and comprehensive approach to measurement of the quality of healthcare, performance measurement will ultimately fail to achieve its quality improvement goals.

A technical solution to improving palliative and hospice care

PurposeThis project sought to help palliative and hospice care practices improve patient care quality and operational efficiency by improving patient symptom status reporting and symptom management, reducing associated provider documentation workload, and enhancing patient–provider and provider–provider communication. We developed a user-friendly, electronic medical record-compatible, software prototype that allows typical clinical data and patient-reported outcomes (PRO) to be entered and stored. This data is immediately available during the clinical encounter with graphically depicted summaries for patient history and PRO assessments, a trending feature that links symptom behavior to interventions and the Edmonton Labeled Visual Information System.MethodsA user-centered design approach allowed for iterative cycles of needs/usability feedback from providers and patients/caregivers to be incorporated into the development of our prototype’s technical structure and features. To determine the needs and initial usability of the project’s prototype, we interviewed eight providers and 18 patients/caregivers. Another usability test, consisting of patient/caregiver (n = 18) and provider (n = 9) interviews, assessed the functioning prototype’s design, usability, and usefulness.ResultsPatients/caregivers (n = 18) reported that the prototype was usable (100%), it would facilitate patient–provider communication, shared decision making, and self-management (100%), and they would be willing to try the system and recommend it to their providers (100%). The providers (n = 9) felt that the prototype encouraged better use of patient assessments in decision making and patient care (100%) and improved identification of cause/temporal relationship between care events and outcomes (100%), monitoring of patient status (100%), communication in a multi-disciplinary team (100%), and operational efficiency and patient care quality (88.9%).ConclusionsQuality of patient care and operational efficiency can be improved with an effective assessment, evaluation, and communication tool. This project developed an electronic version of such a tool. Future efforts will hone its usability and integration across multiple hospice/palliative care settings.

The symptom burden index: Development and initial findings from use with patients with systemic sclerosis

Objective. Our study had 3 aims: (1) to evaluate the functioning of the Symptom Burden Index (SBI) in patients with systemic sclerosis (SSc); (2) to determine the amount of burden per problem experienced by patients as well as the number of patients experiencing each measured problem area, and the number of SSc problems per patient; and (3) to characterize the burden profiles of problem area-specific subgroups of patients. Methods. We developed the SBI to determine the effect of problems in 8 major symptomatic areas of importance to patients (skin, hand mobility, calcinosis, shortness of breath, eating, bowel, sleep, and pain). Results. Sixty-two patients with SSc completed questionnaires on current disease-related problems, physical functioning, and health status. On average, patients were 53.4 years old and had had SSc for 8 years. Patients were mainly women (87%), English-speaking (87%), with diffuse SSc (63%), white (69%), married (61%), and lived with 1 or more additional household members (84%). Only 26% were employed full-time. The 3 most widely reported problem areas were pain, hand, and skin, experienced by 92%, 89%, and 88%, respectively. About one-third reported experiencing 0–5 problems and one-third 7–8 problems; individual patients experienced, on average, 5.7 problems. Conclusion. Psychometric evaluation determined that (1) summarizing SBI problem area item sets to report burden scores per problem measured is justified; (2) the 8 proposed problem areas are independent and deserve separate evaluation; and (3) burden scores correlate as expected with the Health Assessment Questionnaire-Disability Index and the Medical Outcomes Study Short-Form 36 questionnaire. The number of problems experienced and the degree of problem-associated burden that patients with SSc bear are substantial. Use of the SBI’s patient-focused measurements may aid physicians in resolving problems most directly affecting patients’ quality of life. This approach to measuring symptomatic burden in patients with chronic disease could be extended to other conditions.

Effect of social support on informed consent in older adults with Parkinson disease and their caregivers

PURPOSE: To evaluate the effects of social support on comprehension and recall of consent form information in a study of Parkinson disease patients and their caregivers. DESIGN and METHODS: Comparison of comprehension and recall outcomes among participants who read and signed the consent form accompanied by a family member/friend versus those of participants who read and signed the consent form unaccompanied. Comprehension and recall of consent form information were measured at one week and one month respectively, using Part A of the Quality of Informed Consent Questionnaire (QuIC). RESULTS: The mean age of the sample of 143 participants was 71 years (SD = 8.6 years). Analysis of covariance was used to compare QuIC scores between the intervention group (n = 70) and control group (n = 73). In the 1-week model, no statistically significant intervention effect was found (p = 0.860). However, the intervention status by patient status interaction was statistically significant (p = 0.012). In the 1-month model, no statistically significant intervention effect was found (p = 0.480). Again, however, the intervention status by patient status interaction was statistically significant (p = 0.040). At both time periods, intervention group patients scored higher (better) on the QuIC than did intervention group caregivers, and control group patients scored lower (worse) on the QuIC than did control group caregivers. IMPLICATIONS: Social support played a significant role in enhancing comprehension and recall of consent form information among patients.