Development of the Self-Administered Health Complexity Screening Instrument

The purpose of the study was to develop a self-administered screening tool to assist case managers, primary care physicians, and other clinicians in quickly and accurately identifying patients with complex health needs who require a definitive and detailed case management evaluation. In addition to medical and psychiatric items, this tool incorporates multiple social determinants of health (SDOH), known to add to clinical complexity.

This instrument is most appropriate for identifying patients with high complexity in the primary care sector. Due to the abundance of such patients with high biopsychosocial complexity and costs of treating these patients, this tool can select those patients at greatest need for more detailed and comprehensive evaluation and treatment planning to identify and address the patient's barriers to health improvement.

Potential participants were recruited from medical, psychiatric, and addiction outpatient and inpatient settings at a large metropolitan medical center. In Phase 1 of the study, a pool of potential screening items was drawn from several sources: the self-rated and interviewer-rated Patient Centered Assessment Method (PCAM; Pratt et al., 2015 ), the INTERMED self-assessment instrument (IMSA; Van Reedt Dortland et al., 2017), and additional items generated by one of the authors (S.A.F.). Forty participants completed this phase. Statistical analyses yielded a homogeneous scale with a smaller number of items. The screening instrument (Health Complexity Screener [HCS]) was then correlated with the criterion measure, the Value-Based Integrated Case Management Complexity Assessment Grid (VB-ICM-CAG; Kathol et al., 2018 ) with 35 participants, resulting in a receiver operating characteristic (ROC) analysis.

Sixteen items were selected from a larger item pool based on item total correlations. This instrument was then correlated with the gold standard VB-ICM-CAG, and the internal consistency reliability was assessed (standardized coefficient alpha of .892). The VB-ICM-CAG criterion measure was dichotomized at the median value. An ROC curve analysis produced a cut score on the Screener that best represented the tradeoff between an elevated proportion of true positive (TP) rates above the median and a low proportion of false positive (FP) rates. The summary finding of predictive accuracy or area under the curve (AUC) was .83 ( p = .001), with a TP rate of .81 and a FP rate of .21. These data suggest a screening instrument with acceptable psychometric characteristics.

For patients of Primary Care Clinics, Mental Health Clinics, and/or Behavioral Health Providers, the HCS is an efficient way to screen all new patients, quickly identifying those who need additional case management assessment and/or assistance to follow their provider's treatment plan. For patients in facility care (e.g., inpatient, residential treatment facility, and skilled nursing facility), this screening tool would augment the discharge planning process by quickly identifying patients at high risk of readmission due to barriers to following the discharge plans. For insurance case managers, this tool would be an efficient way for intake coordinators to effectively screen and identify members for complex case management versus those that would benefit from less intensive case management services.