Purpose: To explore the real-world utilization of computerized tomography (CT) in patients with advanced colorectal cancer (CRC) and the associated outcomes. Methods: Using Optum's de-identified Clinformatics (R) Data Mart Database (2008-2016), we identified patients with CRC receiving combination of chemotherapies (fluoropyrimidines with either oxaliplatin or irinotecan, or capecitabine with either oxaliplatin or irinotecan) combined with bevacizumab as the initial treatment, and its starting date was registered as the index date. End of treatment was defined by the presence of a gap in therapy > 60 days or treatment switch. We assessed the CT scan utilization during the period between 60 days pre-index and 30 days post-end of treatment. Cox regressions were performed to assess the impact of intensity of follow-up CT scans, measured by time to the first scan from the index date, on likelihood of treatment switch and survival. Results: Among included 4,810 patients, the median (standard deviation) time to the first follow-up scan was 57 (45) days. The mean and median number of CT scans was 4 and 3, respectively. An earlier follow-up scan was associated with significantly greater chance of treatment switch (hazard ratio = 0.99 for each day of delay, P < 0.01), and greater likelihood of death (hazard ratio = 0.99 for each day of delay, P < 0.01). Conclusion: : More intense follow-up increased the likelihood of treatment switch, yet it was not associated with better survival outcome. Further analysis in populations with longer follow-up period is warranted to elucidate the link between CT scan utilization and outcomes.

Therapeutic inertia is the failure of health-care providers to initiate or intensify therapy when therapeutic goals are not reached. Few studies have compared insulin initiation delay times with their expected negative effect on health outcomes. Our study quantifies the effects of delaying the start of insulin for patients with Type 2 Diabetes Mellitus (T2DM) on future A1c values, medical costs, and development of new T2D-related complications.

We aimed to explore the real-world utilization of Computerized Tomography (CT) in patients with advanced colorectal cancer (CRC) and investigate whether a more intense use of follow-up CT scan improves patient outcomes.

Objectives: Evaluate the impact of pharmacist-provided transition of care (TOC) services on hospital readmissions. Methods: Starting March 2014, TOC services were provided to all hospitalized patients from an at-risk medical group. Data covering all inpatient and outpatient services and prescription drugs were retrieved for all adult patients discharged between January 2010 and December 2018. The overall impact of TOC was estimated using a generalized estimating equation with logistic regression. Longitudinal TOC effects were estimated using generalized estimating equation in an interrupted time series model. Parallel analyses were conducted using data from an affiliated medical group in a neighboring county without access to the TOC intervention. Results: The study included 13,256 hospital discharges for adult patients for the 30-day readmission analysis and 10,740 discharges for the 180 days analysis. The TOC program reduced 30-day readmission risk by 34.9% [odds ratio (OR)=0.651 (range, 0.590-0.719)] and 180-day readmissions by 33.4% [OR=0.666 (range, 0.604-0.735)]. The interrupted time series results found the 30-day readmission rate to be stable over the pre-TOC period (OR=0.00; not significant) then to decreased by 1.5% per month in the post-TOC period [OR=0.985 (range, 0.980-0.991)]. For 180-day readmissions, risk decreased by 1% per month after TOC implementation [OR=0.990 (range, 0.984-0.996)]. Referral to the medical group's pre-existing Priority Care clinic also reduced readmission risk. Results from the comparison medical group found 180-day readmission declined by 1% per month after March 2014 [OR=0.990 (0.891-1.00)]. Conclusions: Adding a pharmacist-led TOC program to the medical group's existing outpatient services reduced 30- and 180-day readmissions by "bending the curve" for readmission risk over time.

Persistence with initial antidiabetic treatment regimens is critical for reducing the risk of adverse clinical events and costs. This study estimated the impact of persistence on clinical and cost outcomes among patients with type 2 diabetes (T2D). We identified adults aged over 26 years with T2D who initiated treatment with oral antidiabetic therapy using Optum’s De-identified Clinformatics® Data Mart database [2007-2018]. Patients were required to have continuous enrollment >=12 months before and >=12 months after the index prescription. Persistence was measured using duration of initial therapy, which was calculated as the days between the index date and end of the day’s supply prior to discontinuation [indicated by a gap in all therapy >=60 days] or last available fill date, whichever occurred first. The associations between persistence and the risk for stroke, acute myocardial infarction (AMI), and all-cause or cardiovascular disease-related hospitalizations were estimated using time-varying multivariable Cox proportional hazards models to capture the temporal relationship between discontinuation and adverse events. The impact of persistence on health care costs was estimated using generalized linear models (GLM). A total of 229,485 patients met study selection criteria. Among them, 41.5% of patients discontinued their initial therapy. One month increase in the duration of initial therapy was associated with significant reductions in cardiovascular event risk (6%), all-cause hospitalization (2.4%), and cardiovascular disease-related hospitalization (1.6%). One month increase in the duration significantly decreased total healthcare costs per patient in the first year by $12.15 per month. This effect consisted of significant decreases in medical costs ($12.48), outpatient costs ($6.43), and inpatient costs ($15.88) which were partially offset by a small but non-significant increase prescription drug costs ($0.91, p=0.078). Patients who are persistence on their initial antidiabetic medications enjoyed significantly reduced risks of adverse health outcomes and lower healthcare costs.

Document the predictors of duration on initial therapy, following a diagnosis, measured as days of continuous therapy. A retrospective analysis was conducted using a 1% data subset from a large private health insurance claims database, years 2008-2017. Subjects were included based on having an initial hypertension diagnosis and filling an antihypertensive prescription (N=26,917). The date of first therapeutic fill is a index date. Subjects were required to have at least 6 months of data in the pre-index and 12 months post-index. The covariates of interest included the class of antihypertensive used as initial therapy. The multivariable analyses controlled for age, gender, health insurance type, geographic region, prior comorbidities, and prior 6 months health expenditures.Analyses on duration utilized ordinary least squares (OLS). Analyses for time to event outcomes utilized Cox survival models. The study sample 47.5% male, generally on an HMO or POS, and an average age of 64.4. ACE inhibitors were the most common initial therapy (21.5%), followed by Beta Blockers (19.7%), Calcium Channel Blockers (14.9%), and Angiotensin Receptor Antagonists (10.2%). The average duration of initial therapy was 66.9 days (sd=29.4). We did not find statistically significant difference between the class of medications used. Other results found that age and state are currently the best predictors of duration of initial therapy. Time to discontinuation showed similar results, using Cox models, with age and state currently being the only significant covariates. This was a preliminary analysis based on a 1% sample of Optum patients. Work in underway to re-create these data using the full Optum sample and further investigation is underway.

Lack of medication persistence with antidiabetic drug therapy increases risk of diabetes complications and hospitalizations. This study assessed the impact of type of initial antidiabetic medications, demographic and clinical factors on persistence among adult patients with type 2 diabetes (T2D). We identified adults aged over 26 years with T2D who initiated treatment with oral antidiabetic therapy using Optum’s De-identified Clinformatics® Data Mart database [2007-2018]. Patients were required to have continuous enrollment >=12 months before and >=12 months after the index prescription. The initial treatment regimen was defined as all medications used in the 6-week period following the first antidiabetic medication to capture adjustments that are common in real world practice. Persistence was measured using time to discontinuation [time to a gap in all therapy >=60 days]. The primary independent variable of interest was the medication class(es) used as initial therapy. Demographic and clinical characteristics included age, gender, health insurance, prior healthcare utilization, diagnostic mix and prior use of non-diabetic medications. The association between initial therapy and patients’ characteristics and the likelihood of discontinuation was estimated using Cox proportional hazards models. A total of 229,485 patients meeting study criteria were identified. Patients initiating treatment using sulfonylurea monotherapy, thiazolidinedione (TZD) monotherapy, or combination therapy were 5.2%, 28.7%, and 19.9% significantly less likely to discontinue initial therapy than patients initiating treatment on metformin monotherapy. Patients who were older, had lower prior medical costs, and higher prior drug costs were significantly less likely to discontinue their initial therapy. Patients who received anti-infectives, autonomic drugs, central nervous agents, muscle relaxants, and respiratory agents were significantly more likely to discontinue while patients who received cardiovascular drugs and electrolytic agents were significantly less likely to discontinue. Patients initiating sulfonylurea, TZD, or combination therapy were significantly less likely to discontinue their initial treatment regimen.

To evaluate the impact of pharmacist-provided transition of care (TOC) services on 30-day and 180-day hospital readmissions for patients of the Bakersfield Family Medical Center (BFMC), an at-risk physician group. Synergy Pharmacy Solutions (SPS) provided TOC services for all BFMC hospital discharges beginning in March 2014. The dataset included claims for all patients hospitalized between January 2010 and December 2018, including demographics, inpatient and outpatient visits. All discharge episodes were screened for 6 months of continuous enrollment prior to the index hospitalization and a 30 days of post-discharge data. A second screen requiring 180 days of post-discharge data was applied for the 180-day analysis. The average TOC intervention impact was estimated using multivariable logistic regression, and TOC effects over time were estimated with an interrupted time series (ITS) model. The study included 13,256 hospital discharge episodes for the 30-day readmission analysis and 10,740 discharge episodes for the 180 days analysis. The logistic regression model estimated that the TOC program reduced 30-day readmission risk by 19.3% [OR=0.807, p<0.0001] and reduced 180-day readmissions by 18.4% [OR=0.816, p<0.0001]. In the ITS model, the 30-day readmission rate was increasing by 0.2% per month [p=0.3517] prior to TOC implementation. After implementation, the trend was reversed, and the risk of 30-day readmission decreased by 1.5% per month [p<0.0001]. For 180-day readmissions, the ITS model showed that risk did not change over time prior to TOC implementation, then it began decreasing by 1% per month after TOC implementation [p=0.0012]. Readmission risk was also significantly reduced for those referred to BFMC’s outpatient clinic for management of chronic diseases. Adding a pharmacist-led TOC program to BFMC’s outpatient services significantly reduced 30-day and 180-day readmissions and “bent the curve,” reducing readmissions over time.

e14011 Background: Chemotherapy combined with bevacizumab is the most commonly used treatment first line therapy in pts with mCRC. Decisions for continuation or change of chemotherapy are based on the findings of CT scan, the most commonly used from of imaging in this population. Modeled after clinical trials, CT scan every 2 months is adopted as a standard of care. Yet, patterns of utilization of CT scan in general population is unknown. We set to explore CT scan utilization and associated outcomes among pts with mCRC. Methods: The De-identified Clinformatics Data Mart (OptumInsight, Eden Prairie, MN) covering January 2008 to December 2016 was used for this analysis. Pts with two out-patient and/or one in-patient ICD codes for colorectal cancer were identified. Pts with at least 180 days of enrollment, no chemotherapy within 120 days prior to chemotherapy, and at least one claim for CT scan were eligible for analysis. Recipients of FOLFOX (CAPOX) or FOLFIRI (XILIRI) +bevacizumab were identified using HCPCS codes and the data of their 1sttreatment was registered as index date. The primary endpoint of the analysis was exposure to both FOLFIRI and FOLFOX, secondary endpoint was survival. SAS software was used for data processing and analysis. Results: A total of 3261 pts met the inclusion criteria 78% with oxaliplatin based regimens and 22% with irinotecan regimens. The median age of the population is 66 (19-89), and 58.3% of the identified pts were male. The median duration of first line therapy was 119 days. Median number of CT scan during first line was 2.3. The median and mean number of CT scan per 2 months were 0.82 and 0.94. There was no difference in age, gender, and comorbidities in those with less than 2 vs. 2 or more CT scans. Exposure to both regimens (measured with switching from one regimen to another) was 35% in pts with less than 2 CT scans and 44% in those with 2 or more CT scan (p-value < 0.0001). Probability of survival at 12 months was 83% for all patients regardless of the frequency of scans. Conclusions: In patients with mCRC more frequent scans is associated with higher probability of access to active agents. However, survival probability at 12 months was not different between the two groups.

Monitoring disease status and treatment planning for mCRC is contingent upon CT scan findings. Yet, due to limitations of case identification in the claims data, the frequency of CT scan in patients with mCRC in the real world setting and its impact on treatment duration is unknown. We devised a method to explored the utilization of CT scan and treatment pattern among mCRC patients who received National Comprehensive Cancer Network (NCCN) recommended first-line chemotherapies in combination with bevacizumab. The De-identified Clinformatics® Data Mart (OptumInsight, Eden Prairie, MN) covering January 2008 to December 2016 was used for this exploratory analysis. Patients with mCRC and at least 180-day continuous insurance enrollment were identified using ICD codes. Receients of FOLFOX or FOLFIRI +bevacizumab were identified using HCPCS codes and the data of their 1st treatment was registered as index date. SAS® software were used for data processing and analysis. A total of 1,120 colorectal cancer patients in the 1% sample of the OPTUM enrollees were captured and 32 of them received predefined treatments. Among the identified patients, 24 were FOFLOX recipients and 8 firstly received FOLFIRI; 68.75% were male; the median and mean age were 67.5 (41-84) and 66. The median and mean treatment duration were 153 (12-308 ) and 134.5 days respectively. 46.9% of the patients switched between the two therapies under study. The mean total CT scan frequency is 2.59 (1-9). The mean CT scan frequency per cycle (14 days) is 0.27 (0.08-0.75). The patient population and treatment patterns conform to the previously described treatment patterns in US and thus a valid dataset for exploring frequency of CT scan. We are accessing the 100% sample of OPTUM claims and will apply a regression model to estimate the association between CT scan frequency and treatment duration.

Statins and other lipid-lowering therapies (LLTs) are used to control elevated LDL cholesterol levels and reduce risks of cardiovascular disease (CVD). This study documents the duration of drug therapy achieved during all treatment episodes initiated by patients and estimates how duration of drug therapy impact the risk of CVD events. A retrospective database was created from medical and pharmacy claims covering Jan 1, 2007 to Dec 1, 2017. LLT drug claims were summarized into episodes. An episode was defined when each time a patient initiated treatment, restarted a previously used LLT or switching to a new LLT. Patients were required to have at least 1 year of pre-index and 2 years post-index enrollment data. Drug therapy outcomes included episode duration, MPR, time to and type of subsequent episodes. Clinical outcomes included time to CVD events and all-cause health costs, measured over the 2 years following the index date. Our test sample was drawn form a 1% sample from the overall database. 53% of LTT episode were for primary prevention and 47% were secondary prevention. 54% of patients were female. At index, the mean age was 58 years and the mean LDL was 142 mg/L. Patients had a mean episode duration of 383 days, and only 30% had a 12-month MPR>=80%. 20% of patients had only one episode, 59% restarted their initial therapy, 20% switched therapies, and 2% initiated an augmentation episode. Average time to restart and switch was 124 and 493 days, respectively. 58% patients had a post-index CVD diagnosis and 27% had a post-index hospitalization. Time to CVD diagnosis and hospitalization was on average 556 and 868 days, respectively. Mean of 2-year total health care cost was $10714. LLT compliance remains suboptimal and early treatment termination is common. Further analyses need to be done on subsequent episodes.

While the initial hospitalization accounts for 75% of total healthcare costs during the first 100?days following hematopoietic stem cell transplantation (HSCT), there is a lack of studies evaluating the considerable variation in cost estimates. Using the National Inpatient Sample (NIS) database from 2012?2014, we identified 1832 adult non-Hodgkin lymphoma (NHL) patients who received autologous or allogeneic HSCT and examined complications as predictors of hospital cost. Complications occurred in >70% of patients, and the presence of one or more complications was associated with an increase in mean hospital costs of 46% in autologous HSCT and 81% in allogeneic HSCT. The most common complications (?40%) were mucositis, febrile neutropenia, and infection. Acute organ failure, acute graft-versus-host disease, and death were less frequent (?10%) but had a greater impact on increasing hospital costs and length of stays. Despite recent advances in supportive care and pre-conditioning regimens, complications are common and costly during HSCT.

OBJECTIVES:The 2017 American College of Cardiology/American Heart Association High Blood Pressure Guidelines lowered high blood pressure (BP) threshold, recommending earlier treatment to prevent cardiovascular disease. This study estimated the impact of initiating early antihypertensive medications on the risk of acute myocardial infarction (AMI), stroke, death, and on healthcare costs in patients potentially qualifying for antihypertensive treatment under the 2017 guidelines. METHODS:High-risk patients qualifying for antihypertensive medications under the 2017 guidelines were identified using Optum data. Patients with a diagnosis of elevated BP were also assumed eligible for hypertension treatment under the new guidelines. Patients were defined to have initiated early treatment if they initiated treatment before experiencing a cardiovascular event postdiagnosis. RESULTS:A total of 916 633 patients met eligibility requirements and all other study inclusion criteria. Of those, 66% initiated treatment during 2007-2016. Initiating early antihypertensive treatment decreased the likelihood of having AMI by 59%, stroke by 60% and death by 9%. Patients with only an 'elevated BP' diagnosis experienced reduced risk of stroke once they initiated medications. Treatment reduced the risk of AMI or stroke for patients with diabetes, chronic renal disease and obesity and also significantly lowered all-cause healthcare costs in the first postindex year. CONCLUSION:Initiating antihypertensive medications before experiencing a cardiovascular disease-related clinical event was associated with reduced risk of AMI, stroke and death for all hypertensive patients identified in the new guidelines. However, early treatment had a significantly smaller effect for patients with only 'elevated' BP, who experienced just a lower risk of stroke once treated.

Acute otitis media (AOM) in children is commonly treated with antibiotics, yet limited data of prescribing patterns and the outcomes exists. Document factors affecting antibiotic prescribing patterns for pediatric AOM patients and estimate the effects of antibiotic prescribing on the likelihood of revisits within 28 days. Optum Insight Clinformatics claims data from 2011-2013 were used to identify children (<18 years old) diagnosed with AOM. An episode of care was defined around the first AOM visit (+/- 6 months). Multivariable logistic regression was used to identify demographic and clinical factors associated with antibiotic use. A similar model was used to estimate the effect of treatment on the likelihood of initiating a revisit within 28 days. 2378 (64%) of 3747 children with AOM were treated with an antimicrobial within 28 days, and only 57% of treated patients had their antibiotic prescription filled within 3 days. The likelihood of filling within 3 days increased with age and co-morbid fever or URI, and was lower among patients with a prior respiratory infection. Children receiving the antibiotic within 3 days had a reducedrisk of a revisit within 28 days (adjusted OR=0.439, p<0.001) than those who received the antibiotic in 4-28 days, or who were not treated. Children receiving an antibiotic on the index visit day had 62% lower risk of a revisit (p<0.001) within 28 days than the children who were not treated.Children < 1 year old were 2.4 times more likely to initiate a revisit. Class of antibiotic prescribed did not impact revisit risk. The statistical factors and magnitude were similar for revisits within 7 and 28 days. AOM patients filling an antibiotic prescription on or within 3 days of the index visit are less likely to initiate a return visit within 7 to 28 days.

To examine the effects of adherence to a patient’s initial antihypertensive therapy on health resource utilization and health outcomes. Adherence was measured using the medication possession ratio (MPR). A retrospective analysis was conducted using data from a large private health insurance claims database, years 2007-2014. Subjects were included based on having at least 2 hypertension diagnosis and having filled an antihypertensive prescription (N=595,056). The date of the first fill is the subject’s index date, and subjects were required to have at least 6 months of data in the pre-index and 12 months post-index. MPR was calculated for the duration of therapy. Health related expenditures were calculated for both the pre and post periods, in terms of pharmacy, inpatient, medical, and total costs. Covariates of interest included age, gender, health insurance type, geographic region, comorbidities, and previous 6 months total health expenditures. Analyses on costs utilized generalized linear models (GLM) and 2-part regression models. Analyses for time to event outcomes will utilize Cox survival models. Descriptively, 53% of the population is male, 80% of the population is under 65, 58% had a point-of-service health plan, 38% had an MPR of 1, and 14% an MPR less than 0.05. Preliminary analyses indicate that a higher MPR correlates with higher pharmacy costs ($808, P<0.001), but lower medical (-$604, P<0.001), inpatient (-$745, P<0.001), and total health care costs (-$389, P<0.001). Time to event models are currently being finalized. Adherence with antihypertensive drug therapy, as measured by MPR, is associated with lower medical, inpatient, and total health care costs.

Clinical trials have shown that granulocyte colony-stimulating factors (GCSF) prophylaxis significantly reduced the risk of febrile neutropenia (FN) in oncology patients receiving myelosuppressive chemotherapies. This study documents the factors associated with the use of GCSF prophylaxis in breast cancer patients and estimates the impact of GCSF prophylaxis on the risks of FN, other infections, hospitalization, and death. A commercial de-identified, HIPPA compliant paid claims dataset covering January 2007 to September 2016 is used for this analysis. Diagnosis codes were used to identify breast cancer patients and document the occurrence of FN. HCPCS codes were used to locate chemotherapy and GCSF claims. The analysis includes only newly diagnosed breast cancer patients with 180-day continuous insurance enrollment prior to the start of chemotherapy. A logistic regression model is used to estimate the association between patient/chemotherapy characteristics and the use of GCSF prophylaxis. A Cox-regression model is then implemented to test the association between GCSF prophylaxis controlling for the concomitant use of antibiotics, chemotherapy FN risk classification, and other patient characteristics. Sensitivity analyses are carried out to test different model specification. Patients using GCSF prophylaxis are less likely to experience FN (HR=0.24, P <0.01) and death within 60 days of chemotherapy (HR=0.22, P<0.01) during their first session of chemotherapy treatment compared to the non-GCSF prophylaxis cohort. GCSF effects are especially significant among patients using high-risk chemotherapy regimens (HR=0.19, P <0.01). No significant association is found between the timing of prophylaxis onset and its effectiveness reducing FN risk. GCSF prophylaxis had no significant effect on infection-related hospitalizations. GCSF prophylaxis effectively prevents FN events in the 1st treatment session of myelosuppressive chemotherapy among breast cancer patients. Patients using high-FN risk chemotherapy regimens benefit from GCSF prophylaxis the most. Patients using intermediate and low FN risk chemotherapy regimens also received insignificant benefit from GCSF prophylaxis.

BACKGROUND: The American College of Cardiology and American Heart Association (ACC/AHA) issued new cholesterol treatment guidelines in 2013.Two of the groups designated for primary prevention were analyzed: patients with a low-density lipoprotein cholesterol (LDL-C) level ≥ 190 mg per dL and diabetic patients aged 40-75 years.OBJECTIVE: To estimate the effects of primary prevention as specified in the 2013 guidelines on cardiovascular event risk and cost.METHODS: Primary prevention patients were identified using laboratory and diagnostic data for Humana members from 2007 to 2013.Potential study patients were classified into 3 risk groups: elevated LDL-C, diabetes, and elevated LDL-C and diabetes.Patients receiving cholesterol-lowering medications before their index date were excluded.Eligible patients were divided into 2 treatment groups: (1) primary prevention patients who initiated treatment before experiencing any cardiovascular disease (CVD)-related event, and (2) patients who either did not initiate treatment until after experiencing a CVD event or never initiated treatment.The associations between initiating cholesterol-lowering medications for primary prevention and the risk for acute myocardial infarction, stroke, coronary angioplasty, or coronary artery bypass graft surgery were estimated using Cox proportional hazards models.The effect of primary prevention on health care costs was estimated using generalized linear models.RESULTS: 91,066 patients met study selection criteria.Primary prevention rates were the lowest in diabetic patients (35%), who were newly designated for treatment in the 2013 guidelines.Primary prevention rates were higher for patients designated for treatment under earlier guidelines: 65% for patients with elevated LDL-C and 78% for the combined LDL-C and diabetes group.Primary prevention treatment was associated with significant reductions in cardiovascular event risk (up to 37%) and lower total allcause costs (by $673) in the first post-index year.CONCLUSIONS: Initiating cholesterol-lowering medications for primary prevention, as specified in the ACC/AHA 2013 guidelines, for patients with high LDL-C and diabetes is associated with reduced CVD event risks and lower health care costs.

BACKGROUND: Postdischarge medication management services have been shown to reduce the incidence of medication-related problems during the transition from inpatient to outpatient care. A pharmacist-run transition of care (TOC) program has been developed to reduce the unplanned readmissions of a high-risk managed Medicaid population after hospitalization. OBJECTIVE: To estimate the budget impact of adding an outpatient pharmacy-based TOC program to a medical benefit from the payer perspective. METHODS: A budget impact analysis was conducted using a decision-tree model developed in Microsoft Excel. The effect on inpatient and total health care costs from the payer perspective was estimated for the 2-year period following initial hospital discharge. Inputs were based on a total plan population of 240,000 lives, with a high-risk population of 7.5%, of whom 37% were hospitalized and potentially qualified for TOC services, resulting in an eligible population of 6,660 patients. The TOC program was assumed to initially cover 30% of the eligible population, with expansion to 60% over the 2 years. We previously reported that this program reduced the risk of readmission by 32% within 6 months and saved the health plan $2,139 per patient referred to the program, inclusive of program cost, compared with patients receiving usual discharge care. Sensitivity analyses were performed to test the impact of uncertainty of model inputs on the results, with the cost of TOC services ranging from $99 to $2,000 per patient referred. RESULTS: The model showed that the TOC program was cost saving at over $3 per member per month in the first 6 months, which translates to over $25 million in total health care cost savings over 2 years. These results were primarily driven by the estimated reduction in inpatient costs associated with the program, which were estimated at $20 million over the 2 years. Sensitivity analyses illustrated that within all the reasonable ranges of model input parameters, including the upper limit of TOC services set to $2,000 per patient referred, the TOC program resulted in cost savings to the health plan. CONCLUSIONS: The TOC program resulted in potential cost savings of over $25 million to the managed Medicaid plan over a period of 2 years, corresponding to over $4 per member per month. Copyright (C) 2018, Academy of Managed Care Pharmacy. All rights reserved.

HEDIS guidelines for pediatric pharyngitis recommend use of rapid group A strep (GAS) test for patients 3 months to 18 years before prescribing an antibiotic. The effects of these guidelines on revisit rates are unclear. We evaluated the factors determining adherence with HEDIS guidelines for pediatric pharyngitis. We estimated the effects of guideline adherence and antibiotic treatment on the risk of revisits within 28 days. Optum Insight Clinformatics claims data [2011-2013] were used to analyze outpatient visits for patients <18 years old with pharyngitis. An episode of care was defined around the initial pharyngitis visit (+/- 6 months). Four patient groups were defined based on testing and treatment. Logistic regression were used to identify factors impacting guideline adherence, and to estimate the effects of testing and/or treatment on the risk of revisit. Overall, 47% of 24,685 pharyngitis patients received the GAS test and 48% of tested patients received an antibiotic. The four groups identified for analysis here were evenly distributed: untested and untreated [27%], untested and treated [26%], tested and treated [22%], and tested and untreated [24%]. GAS testing was the key factor in reducing the revisit risk. Relative to patients with no test and no treatment, GAS testing followed by antibiotic treatment reduced the likelihood of a revisit by 31% [0.53, 0.71], while testing with no treatment led to a reduction of 47% [0.43, 0.65]. Receipt of an antibiotic treatment, controlling for GAS testing, increased the odds of a revisit by 9% [0.96, 1.24]. Testing for GAS, when combined with the appropriate treatment decision, resulted in a significant decline in revisit risk, whereas antibiotic alone did not. While judicious employment of antibiotics is important in managing infection and curbing complications, use of rapid diagnostic tools is the determining factor in reducing revisits for pediatric pharyngitis.