Update in Outpatient Internal Medicine Robert A Gluckman

Update in Outpatient Internal Medicine Robert A. Gluckman, MD, FACP Regent, American College of Physicians Chief Medical Officer- Providence Health Plans Navy Chapter-American College of Physicians October 8, 2011

Goals Using evidence to promote accountable care Practice variation in cardiac care Using evidence to promote shared decision making PSA screening and treatment of early prostate cancer Using “consumer” technology to improve chronic disease management Systems of care Hypertension Hospital discharge “Potpourri” New agents for stroke prophylaxis in atrial fibrillation CT screening for lung cancer Treatment choices in COPD

Appropriate Use of PCI Multicenter prospective study of patients within the National Cardiovascular Data Registry from 7/1/099/30/10 500, 154 PCI’s 71. 1% acute indications STEMI, NSSTEMI, unstable angina with high risk features 28. 9% non-acute indication 17 member expert panel developed appropriateness criteria Classified as appropriate, uncertain, inappropriate Acute interventions 98. 6% appropriate Inappropriate procedures largely stable patients, > 12 hours from symptom onset JAMA 2011; 306: 53 -61

Appropriate Use of PCI Non-acute PCI 50. 4% appropriate 38% uncertain 13. 4%% had CCS Class III-IV angina 5. 1% documented high risk ischemia on non-invasive testing 2. 7% > 2 meds (83. 5% on 0 -1 med) 11. 6% inappropriate 53. 8% no symptoms 68. 7% documented low risk ischemia on non-invasive testing 42. 3% no meds, 53. 5% 1 med

Optimal Medical Therapy in Patients Undergoing PCI Data collected from National Cardiovascular Data Registry on 467, 211 patients Known CAD, 70% lesion with + stress test or 80% lesion with symptoms Excluded patients with ACS, LM disease or EF <30% OMT defined as anti-platelet agent, beta blocker, and statin unless contraindications Analyzed % patients receiving OMT pre and post publication of the COURAGE Trial. JAMA 2011; 305: 1882 -1889

Medication Prescription Rates Before PCI Before COURAGE 9/1/2005 -3/25/2007 After COURAGE 7/1/2007 -6/30/2009 OMT 43. 5% 44. 7% Anti-platelet agent Beta blocker 88. 8% 88. 3% 62. 0% 63. 1% Statin 62. 4% 63. 0%

Medication Prescription Rates After PCI Before COURAGE 9/1/2005 -3/25/2007 After COURAGE 7/1/2007 -6/30/2009 OMT 63. 5% 66. 0% Anti-platelet agent Beta blocker 99. 0% 99. 2% 74. 0% 75. 9% Statin 82. 7% 84. 7% ACE/ARB 57. 7% 60. 7%

Appropriateness of Diagnostic Angiography Retrospective analysis of 565, 504 patients without previous MI or revascularization from 2005 -2008 undergoing elective coronary angiography JACC 2011; 58: 801 -809

Low CAD Rate Hospital High CAD Rate Hospital Beta Blocker Use 38% 50% Low Framingham Risk 33% 21% High Framingham Risk 14% 21% Atypical CP 45% 27% Stable Angina 33% 42% Positive Stress Test 66% 71% Mean PCI Volume 562/yr 802/yr

Patients’ and Cardiologists’ Perceptions of the Benefits of PCI in stable CAD Survey of 153 patients undergoing elective cardiac catheterization with possible PCI and 27 cardiologists at Baystate Medical Center 77% patients had + stress test 65% patients had hx of angina 41% had angina < once per week 41% had activity limited by angina 77% who received PCI reported angina Cardiologists’ reported angina in 98% of these patients Annals of Int Med 2010; 153: 307 -313

Reasons for performing and beliefs about PCI. Error bars represent 95% CIs. LV = left ventricular; MI = myocardial infarction; OMT = optimal medical therapy; PCI = percutaneous coronary intervention. Rothberg M B et al. Ann Intern Med 2010; 153: 307 -313 © 2010 by American College of Physicians

Cardiac Procedures per 1000 Members by Top 6 Regions in 1 Large Employer Group

Opportunity cost: state government State expenditures (all states), 2006 Nat’l Assoc of State Budget Officers, 2007

Practice Variation in the Use of Cardiac Procedures- Summary The ACC is leading the way for specialty society efforts to reduce inappropriate procedures General internists should be active collaborators rather than passive observers in engaging with specialty colleagues in reducing practice variation Medical staff involvement Collegial physician-physician communication Increased transparency of local and regional utilization patterns is coming and can potentially reduce ineffective or marginally effective utilization

Eplerenone in Patients with CHF and Mild Symptoms 2737 patients with NYHA Class II CHF and EF ≤ 35% Hospitalized within the last 6 months or increased BNP Randomized to eplerenone titrated to 50 mg qd vs. placebo Started 25 mg qd and increased to 50 mg qd at 4 weeks If Cr. Cl 30 -49 ml/min started 25 mg qod and titrated to qd if K+ ≤ 5. 0 Dose decreased if K+ 5. 5 -5. 9 Drug stopped for K+ ≥ 6. 0, remeasured after 72 hours and restarted if K+ < 5. 0 NEJM 2011 ; 364: 11 -21

Eplerenone in Mild CHF Eplerenone Placebo CV death or CHF Hospitalization 18. 3% 25. 9% Overall mortality or CHF Hospitalization 19. 8% 27. 4% Hospitalization for 0. 7% worsening renal function 0. 6%

Aldosterone Antagonists in CHF Observational analysis of 43, 625 patients admitted with CHF and discharged home 242 hospitals participating in the Get With The Guidelines-HF Program 12, 565 patient eligible for aldosterone antagonist therapy EF ≤ 35% Serum potassium ≤ 5. 0 m. Eq/L Serum creatinine ≤ 2. 5 mg/dl in men Serum creatinine ≤ 2. 0 mg/dl in women 34. 5% eligible patients received aldosterone antagonist 10. 55% inappropriate or potentially inappropriate use JAMA 2009 302; 1658 -1665

Aldosterone Antagonists in CHF Patients admitted for CHF may have multiple medication changes Clinician concern about initiating multiple interventions may be a barrier to initiating treatment Dose intensification of multiple medications is generally needed in CHF patients Clinical inertia may be a barrier to optimal care in CHF Need for specialty/PCP role clarification

Population Based Prostate Cancer Screening Trial- Goteberg Trial 20, 000 age 50 -64 randomized to PSA every 2 years vs. no screening Screening discontinued at age 69 Elevated PSA, defined as 2. 5 -3, offered DRE, ultrasound and biopsy Men with negative evaluation re-screened every 2 years and biopsied for repeat PSA elevation Primary endpoint- prostate cancer specific mortality Median follow-up 14 years Lancet Oncol. 2010 Aug; 11(8): 725 -32

Control Invited to screen Attendees Non-attendees Prostate CA diagnosed 7. 2% 11. 4% 13. 8% 3. 9% Low risk 2. 0% 6. 1% 7. 8% 0. 6% Moderate risk 2. 5% 3. 6% 4. 5% 1. 0% High Risk 1. 3% 1. 0% 0. 8% Advanced 0. 9% 0. 3% 0. 2% 0. 5% Low risk: T 1, Gleason score ≤ 6, PSA <10 Moderate risk: T 1 -2, Gleason score ≤ 7, PSA <20 High risk: T 1 -4, Gleason score ≥ 8, PSA <100

# invited to screen to prevent one death= 293 # diagnosed to prevent one death= 12

Co-Morbidity and Mortality Results From the PLCO Trial Men with minimal or no J Clin Oncol 2010 29: 355 -61

Men with ≥ 1 significant co-morbidity Significant co-morbidity examples: CAD, MI, DM, HTN, CVA, BMI > 30, COPD, chronic bronchitis

Determinants and Outcome of Watchful Waiting in Men with Prostate Cancer Health Professionals Follow-up Study Cohort 51, 529 men 3331 diagnosed with prostate cancer 1986 -2007 342 chose watchful waiting 51% remained untreated after 7. 7 years Treatment delayed average of 3. 9 years in those undergoing treatment No difference in prostate cancer death or metastatic disease (about 10 per 1000 pt-years) Age, tumor size, Gleason score, PSA level predicted future treatment

PSA > 20 or Gleason >7, or Stage 3 PSA 10. 1 -20 or Gleason 7, Stage 1 or 2 PSA ≤ 10, Gleason <7 Stage 1 and 2

Radical Prostatectomy vs. Watchful Waiting in Early Prostate Cancer 695 men age < 75, life expectancy > 10 years, T 1 or T 2 tumor randomized to RP or WW Only 12% non-palpable T 1 c tumors at enrollment 5% diagnosed by screening ≈ 75% T 2 tumors > 45% with PSA >10 RP patients received hormonal therapy for recurrence WW patients underwent TURP for obstructive symptoms Median follow-up 12. 8 years NEJM 2011 364: 1708 -117

NCCN Guidelines Active surveillance reasonable option for men with Stage T 1 -2 a, Gleason score ≤ 6, PSA < 10 Active surveillance PSA at least every 6 months DRE at least every 12 months Biopsy 6 months if initial bx. < 10 cores Biopsy 18 months if initial bx. ≥ 10 cores www. nccn. org accessed 10/29/10

Prostate Cancer Screening. Conclusions PSA screening may provide small survival benefit Pre-screening, contamination, biopsy threshold and treatment differences limits data interpretation Survival curves start to diverge at about 10 years Significant benefit may take longer Overdiagnosis is a serious health hazard Over 1 million treated Side effects immediate Impact of overdiagnosis could be dramatically reduced if active surveillance was initial treatment strategy in appropriate patients Shared decision making should emphasize delayed benefit, short term risk, baseline health and option for delayed curative treatment Collaboration with urology community to determine treatment options is essential USPTF Grade D recommendation

Mobile Diabetes Intervention Study Randomly assigned 26 primary care practices to 4 study groups No academic affiliation > 10% patients had diabetes Enrolled 163 commercially insured patients, Hg. BA 1 C ≥ 7. 5% No mental health diagnoses, substance abuse, must have internet access Interventions Usual Care Coach + Patient Portal Coach + Portal +Decision Support Diabetes Care published online 7/25/2011

Mobile Diabetes Intervention Study Usual Care Coaching Portal Decision Support Baseline Hg. BA 1 C 9. 2% 9. 3% 9. 0% 9. 9% 12 month Hg. BA 1 C 8. 5% 7. 7% 7. 9% Mean change - 0. 7% - 1. 6% 1. 2% 1. 9%

Txt 2 stop Trial: Smoking Cessation Support via Mobile Phone Text Messaging 5800 patients aged ≥ 16 willing to attempt to quit smoking within 1 month Participants socioeconomically diverse, 44% left school ≤ age 16 Participants responded to ads via text or online All participants allowed to participate in any other smoking cessation program and all provided helpline numbers Lancet 2011: 378: 49 -55

Txt 2 stop Trial: Smoking Cessation Support via Mobile Phone Text Messaging Intervention group received 5 texts daily x 5 weeks, then 3 texts weekly for 26 weeks Core program of 186 standard messages and 713 personalized messages Messages selected from algorithm based on patient specific data Messages provided motivation and behavior change techniques Control group received text message every 2 weeks thanking them for study participation Biochemically verified 6 month quit rate 10. 7% vs. 4. 9%

Randomized Trial of Depression Follow-Up Care by On Line Messaging 208 patients with newly prescribed antidepressant depression in 9 primary clinics in Group Health Cooperative Patients already enrolled in on line messaging Excluded patients with anti-depressant within 270 days, with bi- polar disorder or any previous mood stabilizer, anti-psychotic Randomized to usual care in primary care setting (including psychotherapy, medication, online messaging) vs. usual care plus nurse based outreach via online messaging Outreach included PHQ-9, medication adherence, side effects, facilitated visits, referral as needed 3 contacts plus additional patient initiated contact Scripted response and algorithm based on questionnaire, tight coordination with PCP Review with supervising psychiatrist, did not provide direct care JGIM 2011 26(7): 698 -704

Randomized Trial of Depression Follow-Up Care by On Line Messaging Intervention Control Adherence 81% 61% Second medication 22% 16% Mental health specialist visit 50% improvement in depression score 32% 31% 55% 41% Very satisfied with depression treatment 53% 33%

Care Transitions Intervention (CTI) Quasi-experimental prospective cohort study of FFS Medicare patients in 6 Rhode Island hospitals Assess generalizability of previous RCT Intervention consisted of health coaching Visit in hospital before discharge Home visit within 3 days Telephone call within 7 -10 days Final telephone call by day 30 Coaches worked 18 -24 hours per week and carried case loads of 12 -15 patients Arch Int Medicine 2011; 171: 1232 -37

Intervention Ensure medications taken matched those prescribed and patients can describe which meds to take and how often Identify health conditions and maintain a personal health record Discuss/practice how to schedule a follow-up appointment Patients make their own appointments Help the patient recognize “red flags” that should prompt a telephone call or urgent appointment with provider Enrolled patients in state-wide HIE

Intervention Group Readmission Rates 12. 7% External Control (eligible for CTI 20% not approached) Internal Control (declined CTI or 18. 6% lost to follow up before home visit) Decline 18. 6% Lost to follow up 18. 7%

Simplified approach to treat HTN 45 Canadian family practices able to enroll 30 patients with uncontrolled hypertension randomized to algorithm or guideline based care Uncontrolled HTN define as ≥ 140/90 or diabetics ≥ 130/80 Mean age 61, 15% diabetics Algorithm consisted of: Step 1 - fixed dose ACE/diuretic or ARB diuretic with up titration i. e. lisinopril 10/12. 5 HCTZ; max dose 20/25 Step 2 calcium channel blocker with up titration Step 3 add α blocker, β blocker, or spironolactone

BP Control at 6 months: Algorithm 64. 7% Guideline 52. 7%

Effectiveness of Home BP Monitoring, Web Based Communication, and Pharmacist Care on BP Control 778 patients with uncontrolled hypertension randomized to Usual care Home BP monitoring and Web services Home BP monitoring, Web services, pharmacist care management Outcomes were changes in BP and % patients controlled 12 month follow-up Goal BP 135/85 JAMA 2008; 299: 2857 -67

Effectiveness of Home BP Monitoring, Web Based Communication, and Pharmacist Care on BP Control Usual Care Systolic BP drop -5. 3 mm Home BP +Web -8. 2 mm Diastolic BP -3. 5 mm -4. 4 mm -7. 2 mm % control 31% 36% 56% # E-Mail contacts # BP meds 2. 4 3. 3 22. 3 1. 69 1. 94 2. 16 No difference in primary care visits Modest decline in specialist visits Home BP +pharmacist -14. 2 mm

Intensive Blood Pressure Control in Diabetic Patients- ACCORD BP 4733 patients with Hg. BA 1 C ≥ 7. 5% Age ≥ 40 -79 with CVD Age ≥ 55 -79 at high risk for CVD Systolic BP 130 -180 taking ≤ 3 BP drugs Creatinine ≤ 1. 5 mg/dl or < 1 gm proteinuria Intensive treatment- Goal BP ≤ 120 mm Hg monthly visits x 4, then q 2 mo Control group- Goal BP ≤ 140 mm Hg visits month 1 and 4, then q 4 mo 1° outcome- non-fatal MI, CVA, CV death Mean follow-up 4. 7 years NEJM 2010 362: 1575 -85

2. 1 meds 3. 4 meds Similar use of ACEI/ARB, thiazides, beta blocker, calcium channel blocker

ACCORD BP Trial Intensive Therapy Events %/yr Standard Therapy Events %/yr Primary Outcome 1. 87 2. 09 (NS) CVA 0. 32 0. 53 (NNT x 5 yr 89) Non-fatal MI 1. 13 1. 28 Death 1. 28 1. 19

ACCORD BP Trial Intensive Therapy Standard Therapy Serious adverse events 3. 3% 1. 27% Dizziness on standing 44. 3% 40. 3% Macroalbuminuria (protein excretion ≥ 300 mg/d 6. 6% 8. 7% Consider periodic monitoring urinary protein for overt nephropathy

INVEST Trial Observational secondary analysis involving 6400 diabetic patients with HTN and CAD Patients randomized to calcium channel blocker vs. beta blocker strategy to achieve BP <130/85 Added ACEI and thiazide Patients categorized Tight control systolic BP <130 Usual control systolic BP 130 -139 Uncontrolled systolic BP ≥ 140 Primary outcome- mortality, non-fatal MI, CVA JAMA 2010: 304; 61 -68

INVEST TRIAL BP < 130 BP 130 -139 BP ≥ 140 Primary Outcome 12. 7% 12. 6% 19. 8% Total Mortality 11. 0% 10. 2% 15. 4% Non-fatal MI 1. 3% 1. 7% 3. 1% Non-fatal CVA 1. 0% 1. 3% 2. 4%

INVEST TRIAL- BP Control in CAD Amer J Med 2010 123: 719 -726

Hypertension A simple algorithm may improve BP control 3 -4 drugs may be necessary Patients with uncontrolled hypertension require therapeutic intensification and efforts to improve adherence Team based care and frequent patient/provider contact increases BP control Carefully consider target BP based on patient characteristics

Apixiban for Atrial Fibrillation. ARISTOTLE Trial 18, 201 patients with atrial fibrillation and one additional risk factor for stroke Randomized to a apixiban (Direct Factor Xa inhibitor) 5 mg po bid vs. warfarin Dose reduced to 2. 5 mg if 2 of the following Age ≥ 80 Weight ≤ 60 kg Creatinine ≤ 1. 5 mg/dl Median duration of follow-up 1. 8 years Primary outcome ischemic/hemorrhagic stroke or systemic embolism Secondary outcome major bleeding, death NEJM 2011; 365: 981 -92

Apixiban Event Rate %/year Warfarin Event Rate %/year Hazard Ratio CVA or 1. 27 Systemic embolism 1. 60 0. 79 Hemorrhagic CVA 0. 24 0. 47 0. 51 Total Mortality 3. 52 3. 94 0. 89 CVA, MI, systemic 4. 85 Embolism or death 5. 45 0. 88 CVA, systemic 6. 13 Embolism, death, or major bleeding 7. 20 0. 85

INR median time in therapeutic range 66% ARR for any bleeding 7. 7%/year

New Anticoagulants in Atrial Fibrillation- Summary Apixiban, dabigitran, and rivoroxaban all reduce hemorrhagic stroke, major bleeding Dabigitran only drug to reduce ischemic stroke Apixiban reduces rate of major bleeding and total mortality Cost an issue, especially for Medicare patients or other coverage with greater cost share for drugs Patients well controlled on warfarin may not require med change Beneficial when difficult to control or monitor in adherent patients. Cannot extrapolate to non-adherent patients due to shorter half life

The National Lung Screening Trial 53, 454 patients age 55 -74 with ≥ 30 pack year smoking history, current smoker or quit within 15 years Randomized to yearly low dose chest CT vs. CXR for 3 years Median follow-up 6. 5 years 18, 146 positive tests in CT group 96. 4% false positive 5, 043 positive tests in CXR group 94. 5% false positive Most diagnostic evaluations were f/u imaging NEJMprocedure 2011; 365: 395 -409 4% CT group underwent a surgical

CT Group Radio logy Group Major complication after diagnostic evaluation 11. 6% N=75 8. 6% N=95 Intermediate complication 14. 6% after diagnostic evaluation N=24 12. 5% N=35

247/100, 000 Patient-years 309/100, 000 Patient-years 6. 7% reduction in total mortality

Conclusions- CT screening for lung cancer In this trial, CT screening resulted in a 20% reduction in lung cancer death and 6. 7% reduction in overall mortality Complications uncommon but higher in screened group Areas of uncertainty Overdiagnosis rate Impact of newer technology (i. e. greater mortality reduction vs. more false positives) Cost-effectiveness Duration and interval of screening Target patient population Authors recommended policy makers wait for additional data

Tiotropium vs. Salmeterol to prevent COPD Exacerbation 7376 patients with moderate to very severe COPD 90% Stage II-III Gold Criteria Patients continued other medications >50% used inhaled corticosteroids, short acting bronchodilators Primary outcome- time to first exacerbation Secondary outcomes Risk of severe exacerbations Exacerbations requiring inhaled steroids and/or antibiotics Discontinuation due to adverse effects Follow-up 1 year NEJM 2011; 364: 1093 -1103

Probability of COPD exacerbation Time to first exacerbation decreased by 42 days

Probability of severe COPD exacerbation