To examine how well self-reported data on health, health behaviors, and healthcare utilization by a sample of privately insured patients predict whether they will incur high healthcare costs the following year. 

 A 2012 mail survey of autoworkers from Chrysler, Ford, and General Motors, with 3983 survey respondents linked to their health insurance claims data for 2012 and 2013. 

 High healthcare costs are defined as being in the 75th percentile or higher of healthcare expenditures. Models that include combinations of claims-based measures of expenditures and morbidity and self-reported measures of health, health behaviors, and healthcare utilization are compared. 

 Claims-based measures of healthcare costs and comorbidity for 2012 were strong predictors of whether a patient would incur high healthcare costs in 2013 (C statistic = 0.78). Self-reported measures of chronic conditions, health status, health behaviors, and hospital use are also good predictors of high healthcare costs. However, even the most comprehensive model that included self-reported measures was not as accurate in predicting high healthcare costs (C statistic = 0.73). 

Efficient targeting of high-cost patients is crucial to the success of innovative care delivery models that attempt to lower costs and improve quality of care through more intensive care management of patients. The results of this study show that in the absence of claims data on prior use and expenditures, patient-reported measures of health status and prior healthcare use are reasonable predictors of future healthcare costs for a privately insured population. 

Identifying patients who are likely to incur high healthcare costs is a crucial goal of innovative care delivery models. Insurance claims or electronic health records are often used to identify high-cost patients, but sometimes they are unavailable. The results of this study show that: 

Self-reported health and healthcare utilization based on survey data can be useful in predicting whether privately insured patients will incur high healthcare costs when claims-based expenditure and health information from prior years is not available. 

Among the survey-based measures, questions on inpatient stays and emergency department visits in the previous year were the strongest predictors of incurring high healthcare costs in the following year.

The shift to value-based payment strategies and healthcare providers assuming greater financial risk for the care of patients has generated considerable interest in risk-adjustment methods that can identify patients who will incur high healthcare costs.

 Such methods often rely on medical claims and electronic health records (EHRs) to predict future health needs and healthcare costs of patients based on their prior diagnoses, utilization, and costs.

About 13% of privately insured individuals change health plans in a given year and about 9% change their usual source of care,

 and this appears to have increased since implementation of the Affordable Care Act (ACA).

 However, claims or medical records data are usually not available to private insurance plans or practitioners when they accept and treat new patients.

Due to the lack of data on health history and prior use, health needs assessments (HNAs) or health risk appraisals (HRAs) are often administered by plans to identify patients’ care needs, although it is largely unknown how useful these tools are in predicting which patients are likely to incur high healthcare costs and therefore may require more intensive care management.

 Among other potential limitations, HNAs and HRAs rely entirely on patient self-reporting of health conditions and medical care use, which may be subject to reporting error.

Research based on the nationally representative Medical Expenditure Panel Survey and the Medicare population shows that common measures of self-reported health status—including the single-item self-rated general health measure, the SF-12 measure, and self-reported chronic conditions&mdash;generally performed as well as risk scores derived from claims or medical records.

 The results of 2 studies of Medicaid enrollees showed that self-reported information on health status, health behavior, prior utilization, and other measures were also good predictors of future use of high healthcare costs.

Whether these results are consistent for individuals with private insurance is unknown. In addition, most studies have not directly compared the predictive power of self-reported health measures with claims-based measures of previous expenditures for the same population. The objective of this paper was to determine how well patient-reported data on health conditions, overall health status, health behaviors, and experiences with the healthcare system predict high health costs in the future. To our knowledge, this is the first study that assesses the ability of a comprehensive set of self-reported measures of health, health behaviors, and healthcare utilization to predict high healthcare costs for a privately insured population, similar to studies that have been conducted for the Medicaid population.

 Also, because the data include both prior-year claims data on health expenditures and self-reported measures for the same time period, the study compares self-reported measures and claims-based measures of expenditures in predicting high healthcare costs.

The data for this study were based on the 2012 Autoworker Health Care Survey, a survey of active and retired hourly wage workers from Chrysler, Ford, and General Motors. The survey was sponsored by the National Institute for Health Care Reform, a nonprofit, nonpartisan organization established by the International Union, United Autoworkers; Chrysler Group, LLC; Ford Motor Company; and General Motors. The total survey sample included 8624 hourly wage workers, retirees younger than 65 years (ie, not eligible for Medicare), and their spouses. The sample was randomly selected, with some oversampling of active workers so that the proportion of active and retired workers in the sample was about evenly split.

The survey questionnaire was administered by mail. A consent form was included that asked respondents for permission to obtain insurance claims data for themselves and their spouses. Survey respondents were asked to sign and return the consent form to the survey firm if they agreed to allow their insurance claims data to be linked to their survey data. There were 3983 survey respondents who provided consent and had claims data that we were able to link to their survey responses. The 

) describes the process for determining the final sample for this study.

The survey response rate was 64%. Among survey responders, claims data were available for 46%, for a combined response rate of 29.4% for the sample with linked claims data. Survey weights used in this analysis adjusted for survey nonresponse and differences between individuals who provided signed consent to access their claims data and those who did not provide consent. With these statistical adjustments, the full survey sample and the linked survey/claims sample were very similar with respect to age, gender, race/ethnicity, education, income, health status, and prevalence of chronic conditions, as is shown in the 

The main dependent variable is a binary indicator of whether the individual had high healthcare costs in 2013 (the year following the survey), defined as being in the 75th percentile or higher with respect to total expenditures for 2013 (about $7000 or higher). Total expenditures include combined costs for hospital care (inpatient, emergency department [ED], outpatient), prescription drugs, office-based physicians, tests, procedures, medical equipment, and other services covered by the health plan.

Two variables from sample individuals’ 2012 claims data were constructed for this analysis, as was a categorical measure reflecting total healthcare expenditures for 2012. Categories include being in the 90th percentile of spending or higher ($16,660 and above), between the 75th and 90th percentiles ($6610 and $16,660), between the 50th and 75th percentiles ($2431 and $6610), and below the 50th percentile ($2431 and under). Diagnoses based on 

International Classification of Diseases, Ninth Revision, Clinical Modification 

codes were used to construct a Charlson comorbidity index.

 This is a commonly used measure of health status based on 22 conditions that has been used in prior research to predict mortality and future healthcare costs.

Measures obtained from the survey questionnaire are used to identify all other independent variables in the analysis (see 

 for more detailed definitions of these variables). These include demographic and socioeconomic characteristics (age, gender, race/ethnicity, educational attainment, family income). The survey also asked about a selected number of chronic conditions that respondents reported had been diagnosed by a physician, including hypertension, heart disease, congestive heart failure, diabetes, chronic obstructive pulmonary disease (COPD), arthritis, depression, high cholesterol, and cancer. Also included were distinct measures of both self-rated physical and mental health, as well as work and activity limitations, needing help with personal care, and requiring special equipment due to health problems.

The survey also asked about health behaviors or lifestyle factors that may be related to healthcare costs, including the amount of physical activity respondents engage in on a weekly basis, whether they are a current smoker, and height and weight (to identify individuals who are obese). Self-reported measures of healthcare utilization are also included, such as having a usual source of care, number of hospital stays, and ED visits.

The analysis assessed: 1) how well self-reported information on health, health behaviors, and healthcare use predict being a high-cost patient in 2013 compared with claims-based measures of health expenditures and comorbidity and 2) which self-reported measures appear to be the most important in predicting whether a patient will be considered high-cost in 2013. Ã¢â‚¬Â©Logistic regression models for the likelihood of being a high-cost patient in 2013 were estimated using combinations of the independent variables described above. One model included only information on 2012 expenditures and comorbidity (based on claims data), as well as age, gender, and race/ethnicity. These results were compared with a series of logistic regressions that included only survey variables. These regressions sequentially added groups of survey variables to assess their relative contribution to predicting high-cost patients. A final model included all of the 2012 claims-based expenditures and survey measures described above.

For each model, we computed a C statistic, which is a common measure used to quantify how well a model predicts an outcome better than chance.

 Values range from 0.5 to 1.0, with a value of 0.5 indicating that the model predicts high-cost patients no better than chance and a value of 1.0 indicating that the model perfectly predicts high-cost patients. Models that have a C statistic of 0.7 or higher are considered to be good predictors of the outcome measure.

 reflects the proportion of explained variation in the dependent variable based on a logistic regression model.

 Values can range from 0 to 1.0, with 0 indicating that the model explains none of the variation in the dependent variable and 1.0 indicating that the model explains all of the variation in the dependent variable.

A third measure is the discrimination slope, which, for each model, reflects the difference in the average predicted values for individuals with high healthcare costs in 2013 and the average predicted values for those who did not have such costs. For the best performing models, sensitivity, specificity, positive predictive values (PPVs), and negative predictive values (NPVs) were computed.

Characteristics of the Sample and Percent With High Healthcare Costs in 2013 

Table 1 shows the weighted characteristics of the sample and the percentage with high healthcare costs in 2013. Autoworkers with high healthcare costs in 2012 also tended to have high costs in 2013. Among those in the 90th percentile of costs for 2012, 54.5% were in the 75th percentile or higher of spending for 2013. Among those who were below the 50th percentile of spending for 2012, only 9.2% had high healthcare costs for 2013.

More than half of autoworkers were between the ages of 56 and 64, with an additional 23.8% between the ages of 50 and 55. Most autoworkers were white or African American, not college graduates, and had annual family incomes less than $100,000. The percentages of high-cost patients were lower for younger workers, African Americans, and those with at least some college.

Autoworkers had high prevalence of self-reported chronic conditions, especially hypertension (43.6%), arthritis (36.9%), depression (23.9%), and high cholesterol (48%). One-third had 3 or more chronic conditions during the survey year. The percentage identified as having high healthcare costs in 2013 was especially high for autoworkers with 3 or more chronic conditions.

Despite a high prevalence of chronic conditions, most autoworkers perceived their physical and mental health as excellent, very good, or good. About 1 in 6 (15.6%) reported that they were limited in their ability to work due to health, and 13% reported other health-related activity limitations. High healthcare costs in 2013 were strongly associated with these measures of health status.

About one-fifth of autoworkers reported little or no physical activity in a typical week, 23.2% were smokers, and 42% were classified as obese. The percentage with high costs in 2013 was higher among those who had no physical activity (compared with individuals who were physically active) and were smokers (compared with nonsmokers), and obese persons (compared with the nonobese workers).

Most autoworkers (91.3%) had a usual source of care, 12.5% reported an inpatient stay in 2012, and 7.4% reported 2 or more visits to the ED in 2012. Autoworkers with self-reported inpatient and ED use were much more likely to have high healthcare costs in 2013 compared with those with no self-reported hospital use.

 compares the performances of different models predicting high-cost patients for 2013. Model 1 includes 2012 expenditures, the comorbidity index, age, gender, and race/ethnicity. This model had a C statistic of 0.78, a pseudo 

of 17.8%, and a discrimination slope of .209. Models 2 to 5 include only patient-reported information from the surveys. Model 2 includes patient-reported demographics, education, and income; model 3 adds self-reported chronic conditions; model 4 adds health status and health behavior measures; and model 5 adds self-reported inpatient and ED use. The results for the C statistics, 

, and discrimination slope are consistent in that they show: 1) adding self-reported health, health behaviors, and utilization (models 3-5) substantially improves predictions of high healthcare costs compared with the model that includes only demographics and socioeconomic status, and 2) models based on survey measures have high predictive power (model 5, C statistics = 0.73), but not quite as high as the model that includes claims-based measures and demographic characteristics (model 1, C statistic = 0.78).

 shows measures of sensitivity, specificity, PPV, and NPV for the 3 best-performing models (models 1, 5, and 6), computed at both the 50th and 75th percent risk thresholds. The most noteworthy finding from these results is that measures of sensitivity (ie, the percentage of individuals who had high costs in 2013 who were accurately predicted by the models) are low, relative to similar studies of the Medicaid population.

 In fact, few high-cost cases for 2013 were predicted accurately based on the 75th percentile threshold. Models that include claims data (models 1 and 6) perform better on sensitivity compared with the model that includes only survey data (model 5). The model that includes both claims-based and survey variables performed the best on sensitivity at the 75th percentile risk threshold.

Importance of Individual Self-Reported Health Measures in Predicting High-Cost Patients

 shows the marginal probabilities computed from the model with survey-only variables (model 5). The probability of being a high-cost patient in 2013 was significantly higher among older (relative to younger) individuals, as well as among those with diabetes, COPD, arthritis, depression, and cancer; among those in good, fair, or poor self-reported health (compared with excellent or very good); and among those with work limitations due to health. Those with self-reported hospital stays and ED visits in 2012 were more likely to have high costs in 2013 compared with those with no hospital use.

The results of this analysis show that self-reported information on health, health behaviors, and healthcare use commonly obtained through HNAs or HRAs is a reasonably good predictor of future healthcare costs for a privately insured population in the absence of claims or EHRs. Although the models with only self-reported measures do not perform quite as well as models that include claims-based information on spending and morbidity, the results are similar to studies that examined the usefulness of self-reported measures in predicting high-cost patients among Medicaid beneficiaries.

Although error in patient-reported data is a longstanding concern, one advantage is that it is less susceptible than claims or EHR data to “upcoding,” or the tendency by some plans and providers to aggressively code patient diagnoses to make patients appear sicker in order to maximize payment. A recent study found that risk-adjustment scores based on claims data were significantly higher for enrollees in Medicare Advantage health plans—which are compensated by the federal government and partially based on risk scores&mdash;than they would be if the enrollees were in fee-for-service plans.

 Similar risk-adjustment methods are used in the ACA’s federal and state marketplaces: higher rates are paid to plans with sicker enrollees, funded in part through lower rates paid to plans with healthier enrollees. Self-reported health information from patients or plan enrollees is less susceptible to such bias.

Nevertheless, there are no perfect predictors of which patients will be high-cost in the future. Only about half of the study sample who had high healthcare costs in 2012 also had high healthcare costs in 2013, based on the definition used in this study. This may reflect greater variability among the sample for the study compared with other studies, either due to the relatively small sample size (n = 3983) or because the 25% of the study sample with high healthcare costs was more heterogeneous than a similarly defined group for the Medicaid population. Regardless, the best measures in this study used to predict high-cost patients still have a relatively high rate of error.

The success of innovative care delivery models that focus on care management for high-cost patients depends, in part, on whether the additional resources needed for more intensive care management results in greater cost savings in the long run by preventing unnecessary or avoidable utilization. The key to this success is the efficient targeting of patients who will incur high healthcare costs unless diverted into care management programs. If such targeting includes a large number of patients who will not incur high costs even without the intervention, the effectiveness of care management practices in reducing healthcare costs may be greatly diminished. The high scores for specificity in this study (ie, the proportion of non—high-cost cases in 2013 that were accurately identified as such) suggest the models estimated in this study would be relatively successful in preventing costly case management or other specialized services to patients who would not benefit from them. On the other hand, the relatively low sensitivity scores suggest that a large percentage of patients who would potentially benefit from these services may not be selected to receive them and therefore would be at higher risk for incurring higher costs.

There are several limitations to this analysis that should be noted. First, the sample is limited to US autoworkers and therefore may not be generalizable to other privately insured populations. Predicting high costs for the autoworker population, which tends to be older and have a high prevalence of chronic conditions, may be quite different than for a younger and healthier population. In addition, the small sample size, compared with those in other studies that examined self-reported measures, may lead to lower precision in the predictive ability of self-reported measures than if a larger sample had been available. Also, the results may differ for specific conditions, which is important because many disease management programs are designed to improve quality of care and lower costs for specific diseases (eg, diabetes) rather than high-risk patients in general.

Information from health needs assessments or health risk appraisals are increasingly used for a variety of purposes to improve delivery of care, but little is known as to how effective they could be in targeting privately insured patients who are likely to incur high healthcare costs. The results from this study indicate that self-reported information on health conditions, health status, and healthcare use can be useful in predicting high healthcare costs when prior year claims or medical records are not available.

The following individuals reviewed an earlier version of this draft and provided comments. Written permission has been obtained from all individuals for including them in the acknowledgments. None of these individuals received any compensation for reviewing the manuscript: Paul Ginsburg, PhD, professor of Public Policy, University of Southern California;

Alwyn Cassil, principal, Policy Translation, LLC. In addition, Joel Smith of Mathematica Policy Research, Inc, Washington, DC, provided the programming for statistical analysis. 

 Department of Health Behavior and Policy, School of Medicine, Virginia Commonwealth University (PJC), Richmond, VA.

 National Institute for Health Care Reform.

 The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

 Concept and design; acquisition of data; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important intellectual content; statistical analysis; provision of patients or study materials; obtaining funding; administrative, technical, or logistic support; and supervision.

Articles

WEB EXCLUSIVE

The results of this study show that patient-reported data on health and healthcare can be useful in predicting high-cost patients when claims data for prior years are not available.

Predicting High-Cost Privately Insured Patients Based on Self-Reported Health and Utilization Data

Electronic health records (EHRs) present healthcare delivery systems with scalable, cost-effective opportunities to promote lifestyle programs among patients at high risk for type 2 diabetes, yet little consensus exists on strategies to enhance patient engagement. We explored patient perspectives on program outreach messages containing content tailored to EHR-derived diabetes risk factors—a theory-driven strategy to increase the persuasiveness of health communications. 

Within an integrated healthcare delivery system, women with a history of gestational diabetes participated in 1 of 6 ethnic-specific focus groups to elicit diverse perspectives and a survey yielding quantitative data to contextualize qualitative responses. 

The sample included 35 participants (80% racial/ethnic minorities; mean age = 36 years). Themes regarding tailored messages centered on diabetes risk communication (opposing attitudes about whether to feature diabetes risk factors), privacy (how and whether patient data should be accessed), authenticity (perceiving messages as personalized vs generically computer generated), and preferences for messages sent by one’s personal physician. Trust in the medical profession and perceived risk for diabetes were similar to levels reported in comparable samples. 

 Patient reactions highlight the challenges of leveraging EHRs for tailored messages. Some viewed messages as caring reminders to take preventive action and others raised concerns over intrusiveness. Optimal lifestyle program outreach to improve quality of care for women at high risk for diabetes may require communication from personal physicians, careful development to mitigate concerns over privacy and authenticity, and techniques to counteract the threatening nature of personalized risk communication.  

	Electronic health records (EHRs) present healthcare delivery systems with scalable opportunities to promote preventive lifestyle programs using tailored outreach. We examined perspectives on program outreach messages tailored to EHR-derived diabetes risk factors among women at high risk for type 2 diabetes. 

		Themes from focus group discussions included opposing attitudes about whether to feature diabetes risk factors, how and whether patient data should be accessed, and perceived authenticity of tailored messages. 

		Participants consistently preferred messages sent by a personal physician. 

		Optimal outreach may require communication from personal physicians, addressing concerns over privacy and authenticity and mitigating the threatening nature of personalized risk communication.

Lifestyle programs, or behavioral interventions for weight management, healthy eating, and physical activity, can prevent type 2 diabetes (T2D)

 and are recommended for high-risk adults.

 Such programs are offered by healthcare delivery systems, yet little consensus exists on strategies to enhance patient engagement. Expanding the modest uptake observed in large health systems

 could further increase the programs’ public health impact. These programs are particularly vital for women with a history of gestational diabetes mellitus (GDM). Despite high rates of progression from GDM to T2D,

 and healthcare providers may be unaware of patients’ GDM history due to fragmented care. Such factors potentially influence low program uptake in this high-risk population. Of further concern, considering racial/ethnic disparities in GDM

 minority participation in lifestyle programs appears low

 and minorities appear less likely to use recommended approaches to lifestyle changes.

Electronic health records (EHRs) present opportunities to promote lifestyle programs using tailored messages. In contrast to broad patient outreach, tailored outreach—using messages with content personalized to an individual patient based on demographic, behavioral, and/or theoretical constructs

—is a well-established, theory-based strategy for increasing the persuasiveness,

 of health communications, which may also foster meaningful use of EHRs.

 For example, beyond identifying recipients for a given message, patient-level EHR data could be used to personalize the content of that message (eg, by referencing a specific patient’s combination of disease risk factors).

Automated tailoring has potential as a cost-effective strategy that can be delivered consistently in large health systems, which are key advantages given the potential expense of recruiting eligible patients to available programs.

 Still, concerns about privacy call for caution.

 Furthermore, tailored content in messages inviting participation in a lifestyle program could backfire by linking key aspects of the self to negative disease states,

 resulting in defensive resistance and dismissal of threatening information.

 Patient input is thus critical to inform implementation.

Given gaps in the patient engagement literature, we examined perspectives on tailored outreach messages encouraging women at high risk for T2D to participate in health system-based lifestyle programs. We took a mixed-methods approach to obtain more comprehensive patient perspectives

 and inform future research to disseminate evidence-based lifestyle interventions.

Using a convergent mixed methods design, in 2015, we collected qualitative data via focus groups and quantitative data via survey. Quantitative data served to describe the sample and contextualize focus group responses.

Kaiser Permanente Northern California (KPNC) is a large integrated healthcare delivery system serving 3.7 million members who are demographically similar to the underlying population, except at the extremes of income and education.

The sampling frame consisted of nearly all women diagnosed with GDM in 2011 to 2012 across 44 KPNC medical facilities. These women were previously identified as part of Gestational Diabetes’ Effects on Moms (GEM), a pragmatic cluster randomized clinical trial comparing postpartum T2D prevention strategies. Cluster randomization at the facility level had assigned women to usual care or a 13-session postpartum lifestyle intervention offered as optional routine care on behalf of the health system.

 Here, we used stratified sampling within the GEM cohort to identify women from 6 ethnic groups representative of northern California and among those with the highest prevalence and/or absolute frequencies of GDM.

 Eligibility criteria included being aged 18 to 50 years; comfort reading and speaking English; not currently pregnant; absence of recognized overt diabetes, confirmed by the KPNC diabetes registry

 and self-report; and body mass index (BMI) of 25 to 40 kg/m

 among African American, Mexican American, and non-Hispanic white women and 23 to 40 kg/m

 among Asian Indian, Chinese American, and Filipina women who are at greater risk for diabetes at a lower BMI.

Recruitment included a letter/e-mail followed by telephone invitation and a $40 incentive. Participants provided written or verbal consent, including permission to link data with GEM. Both studies were approved by the KPNC Institutional Review Board.

 Self-reported demographic characteristics, participation in a health system-based lifestyle program in the last 6 months (yes/no), and likelihood of participating in the next 6 months (on a 4-point scale ranging from very likely to very unlikely) were assessed using single items. Patient trust in the medical profession was assessed using a validated 5-item questionnaire; summed responses on a 5-point scale range from 5 to 25, with higher scores indicating greater trust.

 Perceived risk for developing diabetes in the next 10 years was assessed using an item originally derived from the validated Risk Perception Survey for Developing Diabetes (RPS-DD).

 Responses on this 4-point scale range from almost no chance to high chance, with higher scores indicating greater perceived risk. Personal control over developing diabetes was assessed using a 4-item RPS-DD subscale. Averaged responses on a 4-point scale range from 1 to 4, with higher scores indicating greater control. Data from GEM included time elapsed since delivery of the index GDM pregnancy, GEM condition assignment (ie, being in a facility assigned to usual care vs intervention), and whether women in facilities assigned to the GEM intervention had participated in at least 1 session.

 Focus group methods were explicitly chosen to elicit an array of ideas among participants from diverse backgrounds. Rather than interpreting results as confirming consensus or strength of endorsement of specific ideas, qualitative themes can suggest topics for future hypothesis-testing research.

Two researchers co-moderated 6 ethnic-specific focus groups among African American, Asian Indian, Chinese American, Filipina, Mexican American, and non-Hispanic white women. We conducted 1 group in person and 5 via Web-enabled conference calls to maximize participation across a large geographical area. A standardized semi-structured interview protocol using the same moderator prompts

 and predetermined open-ended questions enhanced consistency, neutrality, and comparability of responses across groups.

 Key questions explored outreach message content that would help women decide whether to participate in a lifestyle program and the acceptability of individually tailored messages using EHR data. As a sub-section of the latter topic, participants were asked about the message sender and inclusion of personal diabetes risk factors available in EHRs: history of GDM, blood glucose laboratory results, weight status, and ethnicity (for racial/ethnic minority women). To facilitate discussion, participants were asked about their reactions to a 280-word hypothetical tailored message addressed to a fictional patient that referenced the personal risk factors previously noted; the message was purportedly co-signed by the clinical director of a lifestyle program and department chief of obstetrics/gynecology, and delivered via secure e-mail. Focus groups were digitally recorded and professionally transcribed, verbatim.

Quantitative data were analyzed using SAS 9.3 software (SAS Institute Inc; Cary, North Carolina). We tested for group differences in patient trust, perceived risk for diabetes, and personal control using relevant tests for small sample sizes (eg, Kruskal-Wallis and Fisher’s exact tests). Qualitative data were analyzed using NVivo 10 software (QSR Intl Pty Ltd; Doncaster, Australia) and thematic analysis

 to inductively and deductively derive themes from the data.

 A doctoral-level clinical psychologist and master’s-level researcher reviewed all transcripts to develop a coding scheme of broad meaningful themes. One researcher independently coded transcripts; 2 researchers together then reviewed all coded text and selected key quotes, with discrepancies resolved by consensus. Finally, we integrated quantitative and qualitative data by creating joint displays.

Recruited participants (N = 35) had a mean age of 36 years (standard deviation [SD] = 5.3) and were a mean of 3.6 (SD = 0.3) years postpartum from their index GDM pregnancy. As designed, the sample was ethnically diverse (Table 1). Mean patient trust in the medical profession was 16.6 (SD = 3.8). Regarding perceived risk, 17% (n = 6) believed they had a high chance of developing diabetes, 29% (n = 10) a moderate chance, 49% (n = 17) a slight chance, and 6% (n = 2) almost no chance. Mean personal control over developing diabetes was 3.4 (SD = 0.5). Whereas 6% (n = 2) reported participating in a health system-based lifestyle program in the last 6 months, many endorsed intentions for the next 6 months, with 69% (n = 24) somewhat or very likely and 31% (n = 11) somewhat or very unlikely to participate. There were no significant differences across focus groups in any of the above domains (

 ≥.07). Of 17 women in facilities assigned to the GEM intervention, 88% (n = 15) had participated in at least 1 session; participation was 50% among Filipina women (n = 2/4) and 100% within all other groups.

Six broad themes emerged regarding tailored content in outreach messages, encompassing preferences and concerns. Sample quotes and the focus groups in which each theme emerged appear in Table 2.

 According to one perspective, tailored messages are acceptable and nonintrusive—as long as they are sent securely, by their personal physician, and/or within the boundaries of the health system. As one participant described tensions related to privacy:

“…to sort of pinpoint the people that would most benefit from a [program]…the only way to do that is to get some of that information together about the different people to see who would be the ones that would benefit the most. So, I understand that it’s sort of a really tricky balance between that privacy, but also establishing the access to these programs…it’s a very tricky balance because sometimes those of us who need it most aren’t that quick…to go and get that help.”

Concerns over privacy included wariness about the number of people who might have access to one’s medical records; still, participants acknowledged the advantages of such broad access (eg, the ability to receive integrated care from multiple providers).

 Some participants perceived tailored messages as personal, appealing, and persuasive, creating a sense of authenticity that was interpreted as a genuine expression of professional care—that is, appreciating that “someone took the time to see what my information is.” In contrast, others perceived tailored messages as impersonal, generic, or spam, unlikely to be opened or read; and expressed disdain for ostensibly computer-generated content.

 Participants perceived their physician as trustworthy and concerned about their health and noted that they were more likely to read a tailored message that came directly from that individual. Positive viewpoints also emerged about senders other than a personal physician, founded on the notion that the entire health system is working collaboratively for one’s benefit as a patient. Negative viewpoints about senders other than a personal physician were tied to concerns about privacy, inauthenticity, and perceptions that such messages would lack the caring or potential for follow-up offered by one’s physician.

 Participants described preferences for strong, clear, and hopeful messages that clearly communicated their risk for diabetes and actions needed to prevent it, thus empowering women with the information and resources needed to “get it fixed.” Subthemes included the appeal of messages that recognized women’s past efforts to take good care of their health and a suggestion that messages should convey a sense of urgency. Contrasting perspectives included negative reactions to diabetes risk communication, including taking offense at messages perceived as unhelpful and “telling me what I already know.”

 According to one perspective, inclusion of diabetes risk factors, such as history of GDM, laboratory results, and weight status, was viewed as a caring and persuasive basis upon which to recommend preventive programs. Participants described having “forgotten” that they had a history of GDM and regarded even “unappealing” risk factor information as an important health reminder—particularly in the context of being asymptomatic and engrossed in the competing demands of day-to-day life. A contrasting perspective viewed risk factors as an unwelcome reminder of a physically and emotionally difficult time of life (as standard GDM treatment involves intensive glucose monitoring and control via diet, physical activity, and, occasionally, medication

). Subthemes included feelings of sadness, upset, and taking offense—or imagining that others might take offense, even if they themselves did not&mdash;and preference for messages that focus on the program being offered rather than recipients’ medical histories.

 Perspectives regarding the mention of ethnicity as a diabetes risk factor included seeing it as relevant, useful, and offering a more complete picture of risk factors that acknowledge cultural influences on health, beyond a sole focus on individual choices. Contrasting perspectives included viewing it as uninformative and potentially reflecting cultural stereotypes about health behaviors (eg, negative perceptions of ethnic food preferences).

 In addition to themes relevant to tailored message content, participants preferred that messages specify a range of both program features (eg, cost, staff expertise, behavior change techniques used) and program outcomes (benefits of participating, such as anecdotal and quantitative evidence of effectiveness). The first theme emerged across all focus groups; the second emerged in groups with Asian Indian, Chinese American, Mexican American, and non-Hispanic white women.

Integration of Quantitative and Qualitative Findings

We created a descriptive joint display addressing diabetes risk. The Figure displays the mean proportions of women reporting high perceived risk for diabetes across the focus groups in which risk-related themes emerged. Proportions were somewhat greater across focus groups that expressed preferences for risk information compared with concerns. For example, 19% of women reported high perceived risk in groups that expressed preferences for tailored messages containing clinical risk factors versus 10% in groups that expressed concerns. Patient trust did not markedly differ between groups that expressed preferences versus concerns related to privacy, authenticity, and the trustworthiness of senders (data not shown). For example, trust scores averaged 15.8 across groups that expressed privacy-related preferences and 16.7 across groups that expressed privacy-related concerns.

Although EHR-driven tailored messages have potential as an efficient and cost-effective outreach strategy, divergent patient reactions highlight key challenges. In the present study, diverse women at high risk for T2D raised contrasting ideas about privacy and authenticity. Participants appreciated the size and complexity of integrated health systems and appeared savvy about the ways in which tailored messages could be generated. Yet, whereas some were unconcerned about security, others were strongly critical. Indeed, prior research shows that patients endorse opposing views of what EHR data they want made available to their healthcare team

 and a substantial proportion remain concerned about the privacy of EHRs.

 Seeking input from patient stakeholders and explicitly addressing these issues in tailored messages (eg, by stating how and by whom they have been generated and sent—preferably, by personal physicians) could enhance acceptability and maintain trust.

In terms of risk communication, some participants welcomed it as a caring prompt for preventive action whereas others viewed it as unhelpful, uninformative, or an unpleasant reminder of challenges related to lifestyle behavior change. Similarly, some minority participants appreciated the mention of ethnicity as a diabetes risk factor to acknowledge culture’s impact on health; others raised concerns over stereotyping. These perspectives related to risk factors highlight, first, the complexities of developing sensitive and culturally relevant communications.

 Second, results echo literature on the threatening nature of personally relevant health messages.

 This effect could be counteracted by self-affirmation,

  or asking patients to reflect on a positive aspect of the self—a technique shown to increase attention to diabetes risk information.

 Third, results suggest the need for a positive tone and careful use of clinical information. Patients’ desire for messages that highlight the benefits of lifestyle programs corroborates theoretical

 emphasis on positive consequences and developing favorable expectations (eg, gain-framed messaging) to promote new preventive behaviors. Future research should determine whether theory-based approaches, such as self-affirmation and gain-framed messaging, could mitigate negative reactions to tailored risk communication and increase patient engagement in preventive services. Of note, descriptive findings integrating quantitative and qualitative data raise the possibility that higher self-reported perceived risk may be associated with greater openness to tailored risk communication. If confirmed, health systems could consider surveying patient subsamples to estimate the acceptability of such messages.

Some themes emerged consistently across focus groups (eg, preferences for physician senders), whereas others emerged across fewer groups (eg, concerns about risk communication). Often, the same groups expressed both positive and negative viewpoints signaling within-group heterogeneity (ie, viewpoints differing from person to person) or ambivalence about tailored approaches (ie, the same person expressing conflicting viewpoints), or both. Exceptions include focus groups with Chinese American women, in which themes more often clustered around concerns, and African American women, in which themes more often clustered around preferences. Still, such comparisons must be tentative given that the data came from single focus groups for each racial/ethnic background. Although qualitative focus group methods are well suited to elicit a range of ideas about a phenomenon, identify issues worthy of further investigation, and develop hypotheses, they should be followed by hypothesis-testing research to make definitive cross- and within-group comparisons

Study limitations include the sample’s relatively high level of education, their membership in a single health system, and, as noted, our limited ability to make cross-group comparisons. Strengths include stratified sampling by race/ethnicity and ethnic-specific focus groups to foster a range of diverse perspectives and the mixed methods design, whereby quantitative data provided context in which to interpret qualitative themes. For example, evidence that participants’ trust in the medical profession was similar to a national sample

 and similar between focus groups describing preferences versus concerns about privacy and authenticity suggests that these themes did not arise from undue suspicion. Similarly, perceived risk for, and personal control over, diabetes, reflected levels in comparable samples.

 Of note, the sample was evenly split between women who had and had not been offered a lifestyle intervention in the GEM trial. In responding to focus group prompts, women in the former category may have drawn on their direct experience of having received outreach messages in that context.

In an era of big data, health systems are well poised to discover ways of leveraging increasingly prevalent and powerful EHRs to engage high-risk patients. Results among women at high risk for diabetes suggested that patients acknowledge both the advantages and pitfalls of tailored approaches to outreach. Optimal outreach may require communication from personal physicians, mitigating privacy and authenticity concerns, and applying theory-based approaches to counteract the threatening nature of personalized risk communication.&ensp;

Divergent reactions among women at high risk for diabetes highlight challenges of implementing tailored outreach messages, driven by electronic health records, to promote patient engagement in preventive lifestyle programs.

Leveraging EHRs for Patient Engagement: Perspectives on Tailored Program Outreach

Asthma management programs, such as the Breathmobile program, have been extremely effective in reducing asthma morbidity and increasing disease control; however, their high start-up costs may preclude their implementation in smaller health systems. In this study, we extended validated asthma disease management principles from the Breathmobile program to a smaller clinic system utilizing existing resources and compared clinical outcomes. 

 Cox-regression analyses were conducted to determine the cumulative probability that a new patient entering the program would achieve improved clinical control of asthma with each subsequent visit to the program.

A weekly asthma disease management clinic was initiated in an existing multi-specialty pediatric clinic in collaboration with the Breathmobile program. Existing nursing staff was utilized in conjunction with an asthma specialist provider. Patients were referred from a regional healthcare maintenance organization and patients were evaluated and treated every 2 months. Reduction in emergency department (ED) visits and hospitalizations, and improvements in asthma control were assessed at the end of 1 year. 

A total of 116 patients were enrolled over a period of 1 year. Mean patient age was 6.4 years at the time of their first visit. Patient ethnicity was self-described predominantly as Hispanic or African American. Initial asthma severity for most patients, classified in accordance with national guidelines, was “moderate persistent.” After 1 year of enrollment, there was a 69% and 92% reduction in ED/urgent care visits and hospitalizations, respectively, compared with the year before enrollment. Up to 70% of patients achieved asthma control by the third visit. Thirty-six different patients were seen during 1 year for a total of $15,938.70 in contracted reimbursements. 

A large-scale successful asthma management program can be adapted to a stationary clinic system and achieve comparable results.

A disproportionate health burden exists due to uncontrolled asthma among inner city children.

 Many asthma disease-specific programs have proven to be effective in reducing asthma-related morbidity and improving disease control. Most effective strategies, however, provide long-term follow-up using asthma-specific providers who follow national asthma guidelines.

The Breathmobile/pediatric asthma disease management program (PADMAP) is an effective disease management and treatment program utilizing teams of asthma and allergy specialists who provide regular care to children with asthma in inner cities across the United States via mobile asthma clinics.

 These clinics have consistently achieved impressive asthma-related metrics for the past 20 years, including highly significant reductions in asthma-related hospitalizations, emergency department (ED) visits, and urgent care visits.

 In addition, the programs have shifted urgent care visits to regular care and increased the number of controller medications prescribed.

As with most clinically effective disease management programs, the Breathmobile/PADMAP program requires significant start-up resources. An estimated $365,865 per year was required to purchase and operate 1 Breathmobile for the first 7 years.

 This, however, was recouped in long-term asthma-related health savings during subsequent operating years. In a cost analysis study, the return on investment (ROI) per mobile unit was $6.73 per invested dollar. The annual estimated ED cost in 4 geographic regions was reduced by $2,541,639.

 Therefore, the Breathmobile/PADMAP model is also one of the most cost-effective asthma management programs in the United States. Smaller health systems with smaller budgets and fewer patients may not have the resources to start an effective Breathmobile-like asthma program; yet, these groups have a similar percentage of patients with asthma and, therefore, incur a similar proportion of asthma-associated health and economic burden.

In this study, we describe a collaboration between a tertiary care center with a long established and successful Breathmobile/PADMAP program (Los Angeles County + University of Southern California [LAC+USC] Medical Center) in Los Angeles, California, and a stationary clinic system at Harbor- University of California, Los Angeles (UCLA) Medical Center (HUMC) in Torrance, California, to establish a fixed asthma-specific clinic at the HUMC using previously validated strategies. The need for start-up funds was avoided by using the clinic’s existing resources. We compared the efficiency of settings, asthma-related morbidity, and asthma control to previously published Breathmobile metrics.

 We also determined whether standard Medi-Cal (California’s Medicaid) reimbursements for asthma-related services were sufficient to allow a stationary clinic version of the Breathmobile/PADMAP system to be self-sustaining.

Pediatric patients with asthma were recruited into the study from the Pediatric Allergy-Immunology Clinic at HUMC and the Harbor Medical Foundation (MFI) clinic. These clinics serve the children with Medi-Cal coverage assigned to a large local healthcare maintenance organization (HMO). Patients at the HUMC clinic are referred through the LA County Department of Health Services, EDs, clinics, and inpatient services. Patients at the MFI clinic are referred by HMO asthma disease management coordinators. There were no specific referral guidelines for referring patients to asthma specialty care; however, patients were referred if they were not controlled after being prescribed inhaled corticosteroids, had more than 1 urgent care or ED visit for asthma exacerbation during the last year, or were hospitalized for asthma exacerbation during the last year. All patients had persistent asthma and the majority had previous ED visits and/or were previously hospitalized for asthma exacerbations.

The HUMC/MFI clinics’ operations were modeled identically to that of the Breathmobile/PADMAP system. Staff consisted of 1 allergy—immunology specialist and 2 nursing/asthma educators (1 registered nurse and 1 licensed vocational nurse). One medical technician coordinated appointments and financial matters in a fashion similar to the patient financial worker on the Breathmobile/PADMAP system.

 In contrast to the Breathmobile/PADMAP program which uses full-time allergy—immunology staff, the HUMC/MFI clinic used pediatric nurses who also staffed other pediatric subspecialties, including neurology, nephrology, and rheumatology. Each clinic was held 1 half-day per week.

Patients were evaluated and treated in accordance with National Heart, Lung, and Blood Institute Expert Panel Review 3 (NHLBI EPR-3) asthma guidelines.

 All patients were seen on a regular basis, with follow-up appointments every 6 to 8 weeks. Allergy testing used either skin tests or in vitro—specific-immunoglobulin E (IgE) serum tests (Quest Diagnostics; Madison, New Jersey), as stipulated by the guidelines. Use of in vitro or in vivo testing was left to the discretion of the provider. Generally, patients were allergy tested on their second or third follow-up visit.

The allergy—immunology specialist provider (principal author) had extensive clinical and administrative experience working with the original Breathmobile/PADMAP system. Although the nonprovider staff did not have clinical experience with the Breathmobile/PADMAP system, they were trained by the provider on asthma education in accordance with NHLBI ERP-3 asthma guidelines.

Care coordination and disease management strategies and principles were made in collaboration with the LAC+USC Breathmobile/PADMAP system and transferred to the HUMC/MFI asthma clinic. On an ongoing basis, the allergy—immunology specialist provider had regular meetings with the LAC+USC Breathmobile/PADMAP team to ensure collection of the correct data and execution of the correct strategies.

Disease activity was tracked electronically, similar to Breathmobile/PADMAP programs, in real time and using identical asthma metrics. These included: baseline asthma severity, current disease activity, current disease control, step-up or step-down in asthma therapy, ED visits, hospitalizations, systemic steroid rescue, prescribed asthma controller, Asthma Control Test (ACT) score, and spirometry testing. Breathmobile/PADMAP clinical data is recorded in the AsmaTrax electronic health record (EHR), and our clinics entered identical data fields directly into Excel spreadsheets. Criteria for asthma metrics, such as asthma control, were assessed in accordance with NHLBI ERP-3 guidelines.

In general, patients were given follow-up appointments for 6 to 8 weeks; however, earlier appointments were also scheduled depending on disease severity. Asthma control, asthma-related ED visits, and hospitalizations for patients and medications were reviewed by the provider at the end of each visit. Patients who missed their appointments had these metrics from their previously kept appointment reviewed. An early follow-up visit (2 to 4 weeks) was scheduled for patients who had experienced any of the following conditions since their last visit: 1) more than 3 visits with not well- or very poorly controlled asthma, 2) required ED visit or hospitalization since their last visit, or 3) missed current visit and had not well- or very poorly controlled asthma at their last visit. Patients were given a regular follow-up visit (6 to 8 weeks) if they had all the following conditions: 1) well-controlled asthma at their last visit, 2) fewer than 3 previous visits with asthma control rated as not well- or very poorly controlled, and 3) no ED visits or hospitalizations since their last visit.  Patients with missed appointments who also fulfilled all 3 criteria were scheduled for a return visit within 6 to 8 weeks (

Primary outcomes were reduction in ED visits, hospitalizations, and time to achieve asthma control (asthma rated by NHLBI EPR-3 asthma guidelines as “controlled”). Asthma control, ED visits, and hospitalizations since patient’s last appointment were recorded at every clinical encounter and recorded in EHRs (mobile clinics) and paper charts (fixed clinic). Cumulative ED visits and hospitalizations over a period of 1 year after enrollment were totaled from retrospective EHR and chart review. Time-to-achieving asthma control was defined as the number of visits or provider encounters after enrollment until asthma was rated as controlled. A 1-year period was chosen because the study was funded for this period of time.

The secondary outcome was to determine whether current Medi-Cal/Medicaid reimbursements could sustain operations for patients seen at the MFI clinic, which has a contract with a regional HMO. Patients served as their own historic controls when comparing reductions in ED visits and hospitalizations before and after entry into the program. ROIs with regard to reduced ED visits and hospitalizations were also calculated.

Cox-regression analyses were conducted to determine the cumulative probability that a new patient entering the program would achieve improved clinical control of asthma with each subsequent visit to the program. This incorporated an intent-to-treat approach for each patient, with at least 1 follow-up visit and control status reported at the second visit to the program. To account for the occurrence of multiple events (ie, multiple patients achieving control by a particular visit) and the construct of time as an interval measured as discrete points (ie, achievement of control can occur at any time point between visits), the TIES=EXACT approach was specified in the model statement for the PHREG procedure. This method assumes there is a true but unknown ordering for the tied event times.

 The age of patients, baseline asthma severity, and the 2 sites of care at Harbor-UCLA were controlled for in the analysis.

Patient ages ranged from 3 to 18 years (mean age = 6.4 years; standard deviation [SD] = 4.8 years) at the time of their first visit (

). Patient ethnicity was self-described predominantly as Hispanic or African American. The majority of patients were covered by Medi-Cal and enrolled in HMOs funded by Medi-Cal. Initial asthma severity for most patients—classified in accordance with NHLBI ERP-3 guidelines&mdash;was moderate persistent. Seventy percent of visits occurred within 84 days of the previous visit. The mean interval between patient visits to the program was 64 days. Visit intervals did not differ relative to baseline asthma severity.

Patients engaged in the clinic were likely to achieve asthma control, and those with milder disease were more likely to achieve control sooner. On average, 50% of mild persistent patients with asthma achieved control by the second visit, and by the third visit, 70% of these patients were in control. Patients with moderate or severe persistent asthma took longer to achieve disease control. Approximately 97% of patients with mild persistent asthma, and approximately 90% with moderate/severe persistent asthma achieved control by their sixth visit (data not shown).

Allergy Testing and Environmental Control Education

Patients had allergy testing and environmental avoidance education based on results of the allergy testing. Thirty-three percent of patients received in vitro-specific IgE testing for inhalant allergens; the rest received epicutaneous skin testing. There was no statistical difference between patients who had the 2 different types of IgE testing in terms of probability of achieving asthma control by the third visit (data not shown).

Enrollment in the program resulted in a 69% and 92% reduction in ED/urgent care visits and hospitalizations, respectively. This was achieved by comparing the number of ED/urgent care visits and hospitalizations in the 1-year period following enrollment into the program compared with the previous year before entry into the program (

A total of 116 unique patients were seen at HUMC-UCLA during the 1-year period. Based on reimbursement rates at the time of the study, the costs to deliver care for patients using skin and in vitro testing were $435.50 and $505.70 per patient per year, respectively (

). (These are standard California Medi-Cal reimbursement rates for asthma specialists for high-complexity visits and follow-up.) Overhead for 1 half day a week of clinic services was $10,000 per year.

This study demonstrated that the Breathmobile/PADMAP asthma disease management system could be implemented in a stationary clinic and achieve similar asthma-related outcomes. Existing resources in a multi-specialty pediatric clinic were used to realize treatment goals and were sustained financially from standard Medi-Cal/Medicaid reimbursements.

Patients engaged in the Breathmobile/PADMAP program(s) saw reductions of asthma-related ED visits or hospitalizations from 37.3% before program entry to 8.7% (76% reduction) within the first 6 visits in the program.

 In our clinic, there were more children with ED visits (58%), but similar hospitalizations (24%), in the year prior to program entry. This may be due to the higher number of African American patients within our system. Some evidence suggests that these patients may have poorer response to asthma controllers and relievers compared with patients of other ethnicities.

 PADMAP data also showed that African American children with asthma have a higher risk of losing asthma control even after gaining initial disease control. However, there may be other confounding variables to account for this difference—such as household income&mdash;that we did not examine in this study. The magnitude of morbidity reduction between PADMAP and our system was, however, similar (68% and 91% reductions in ED visits and hospitalizations, respectively, after program entry compared with the year before).

Traditional asthma metrics emphasize reductions in asthma-related urgent care visits and hospitalizations.

 Many patients, however, experience considerable asthma-related impairment from symptoms without requiring urgent care visits or hospitalizations. These impairments include school and work absences, decreased exercise and activity, and reduced quality of life.

 Therefore, a paramount objective of asthma management is to gain asthma control in an expeditious manner. Breathmobile/PADMAP was the first asthma disease management system to demonstrate that regardless of initial asthma severity, up to 70% of children with asthma could achieve asthma control after 3 visits if they remained engaged in the program and were treated in accordance with NHLBI ERP-2 and ERP-3 asthma guidelines.

In our clinic, 70% to 80% of patients achieved asthma control as defined by NHLBI ERP-2 and ERP-3 asthma guidelines after 3 clinic visits. Significantly more patients with an initial diagnosis of mild persistent asthma achieved asthma control compared with those initially diagnosed with moderate or severe persistent asthma by visit 3. However, by visit 6, the probability of asthma control was similar across the 3 groups. Results from Breathmobile/PADMAP and our study highlight the necessity of scheduling regular visits in order to successfully control asthma.

The PADMAP/Breathmobile program is perhaps the most effective ongoing asthma-specific disease-management system in the United States. We chose to extend the services of this successful program to an area where patient demographics were similar. Each Breathmobile is staffed by an asthma specialist provider, 1 RN, 1 respiratory therapist, and a patient financial worker. Since the Breathmobile program’s operational history spanned the publication of 2 editions of national asthma guidelines, care was delivered in accordance with NHLBI ERP-2 and ERP-3 asthma guidelines (dependent on time period).

 We recreated this system within an existing multi-specialty clinic that serves the same inner city children and uses the same diagnosis and treatment guidelines delivered by an asthma specialist. However, due to the smaller size of this clinic, there were not sufficient resources to provide adequate staff for optimal asthma care, and our clinic relied on nurses who also staffed other pediatric subspecialties.

Using existing resources in an established pediatric clinic eliminates the need to raise the large amount of funding necessary to purchase and equip a mobile clinic, which is estimated to be up to $80,000.  A Breathmobile unit may cost up to $241,950 per year to operate for the first 7 years if patients are treated 5 days per week, excluding the salary of the asthma specialist. This translates to $465 per half-day session.

 In our private practice clinic, the total cost—including overhead and excluding compensation for allergy specialists&mdash;was $192 per half day. The operating cost for our system was similar to that of the Breathmobile program. Both systems used allergy specialists and nurse practitioners as providers, and outcomes were similar between PADMAP

 and our clinic (data not shown). As patients in this study were all Medi-Cal—enrolled, we postulate that this successful system can be generalized and adapted to other Medicaid populations in the United States.

In a recent study, the ROI for 4 Breathmobiles in southern California in 2010 was $6.73 saved for each $1 invested.

 Calculations were based on savings resulting from fewer ED visits/hospitalizations, reduced school absenteeism, and quality-adjusted life-years saved. Breathmobiles operated an average of 3.7 days per week, treating almost 600 unique patients per unit to achieve these benchmark outcomes. In our study, we were not able to perform such an extensive cost analysis because we only had data on actual cost to deliver care at 1 of the clinical sites, the MFI clinic, and not the HUMC clinic. However, based upon 2015 reimbursement rates for California Medi-Cal, there was a significant ROI in the context of reduced ED visits and hospitalizations (

 In this study, the cost of care for patients with asthma for a period of 1 year included reimbursements for allergy testing using skin and in vitro allergy testing (Table 2). Both modalities have similar sensitivity and specificity.

 The cost per patient tested with skin testing was slightly lower than similar ones using in vitro tests. Medi-Cal reimbursement rates for allergy skin testing were lower compared with payments for in vitro allergy tests.

 It must be noted, however, that reimbursement rates for both skin prick tests and in vitro allergy tests can vary significantly, depending on the payer. In states such as New Jersey and Illinois, Medicaid payments for skin prick tests and in vitro allergy tests are reimbursed at similar rates.

 In other states, such as North Dakota, and Montana, Medicaid payments for skin prick testing ($9.28 and $6.50, respectively) are reimbursed at higher rates compared with in vitro allergy testing ($6.53 and $3.94, respectively).

 In systems operating with capitated laboratory services arrangements, in vitro allergy testing is customarily included in the per-patient per-month fee, making use of these tests more cost-effective than skin testing, which may require additional payments.

The authors adapted a successful large-scale, specialist-run asthma management program to an existing multi-specialty clinic utilizing existing resources and achieving similar outcomes.

July 2017

Predicting High-Cost Privately Insured Patients Based on Self-Reported Health and Utilization Data

Leveraging EHRs for Patient Engagement: Perspectives on Tailored Program Outreach

Adaptation of an Asthma Management Program to a Small Clinic

Twitter Accounts Followed by Congressional Health Staff

Medication Adherence and Improved Outcomes Among Patients With Type 2 Diabetes

The Price May Not Be Right: The Value of Comparison Shopping for Prescription Drugs

US Internists' Awareness and Use of Overtreatment Guidelines: A National Survey

Availability and Variation of Publicly Reported Prescription Drug Prices

Cost-Effectiveness of a Patient Navigation Program to Improve Cervical Cancer Screening

The Association Between Insurance Type and Cost-Related Delay in Care: A Survey