Richard M. Hoffman, MD, MPHDOIM Thursday SchoolOctober 30, 2014

“But I’m admitting that American medicine has overpromised when it comes to screening. The advantages to screening have been exaggerated.”

Otis Brawley, MD Chief Medical Officer American Cancer Society

New York Times (10/21/09)

Screening definitionCriteria for implementing screeningEvaluating the efficacy of screeningCritical review of prostate cancer

screening Evidence Guidelines

USPSTF cancer screening recommendations

Applying a diagnostic test to asymptomatic people to classify their likelihood of having a particular disease

“Likelihood of disease”

“Likelihood of disease” Persons with abnormal tests

require further diagnostic studies Gold standard tests usually

invasive, riskier, and more expensive

“Asymptomatic”

“Asymptomatic” Patient expectations Screening helpful only if early

detection and treatment is effective▪ First do no harm (primum non nocere)▪ Merely advancing the time of diagnosis

is harmful (lead time)

“Asymptomatic” Target high-risk population▪ Behaviors▪Smoking▪ Risk factors▪Family history

Sporadic (65%–Sporadic (65%–85%)85%)

+ Family+ Familyhistoryhistory(10%–30%)(10%–30%)

HNPCC (5%)HNPCC (5%)FAP (1%)FAP (1%)

Rare Rare syndromesyndromes (<0.1%)s (<0.1%)

CDCCDC

“Asymptomatic” Older age (> 50 years)▪ 70+ million

“Asymptomatic” Expensive▪ Screening▪ Gold-standard diagnosis

“Asymptomatic” Policy decision▪ Limited health care resources▪ Competing demands

Epidemiology

Natural history

Screening tests

Treatments

Important clinical problem Common Substantial burden of suffering▪ Impairs quality of life ▪ Hospitalizations▪ Death

Time course of disease Long asymptomatic period to make

early diagnosis

Acceptable

Available

Accurate: valid and reproducible

Detects clinically important disease

Acceptable

Available

Efficacious Reduces disease-specific mortality

and morbidity in randomized controlled trial

Outcomes Decreased disease mortality and

morbidity▪ Decreased incidence of advanced-

stage disease▪ Prevents disease (sometimes)

Screening study designs Randomized controlled trial: least

biased Observational▪ Cohort▪ Case-control

Selection

Lead-time Overdiagnosis bias

Length-time

Healthy volunteer

Adherent

Worried well

©2006 UpToDate® • www.uptodate.com Licensed to University of New Mexico

Failure to adjust for detecting “pseudo-disease” Subclinical disease that would not

have been detected during the person’s lifetime▪ No chance of dying from the disease▪ Survival time is misleading

Welch HG. JGIM 1997;12:118

©2006 UpToDate® • www.uptodate.com Licensed to University of New Mexico

2014 ACS estimates (men) 233,000 cases (1st) 29,480 deaths (2nd)

Lifetime risks Diagnosis: 1 in 6 Death: 1 in 30

American Cancer Society. Cancer Facts & Figures 2014.

Microscopic cancer (found with PSA testing) to clinical detection 5 to 15 year interval

Prostate-specific antigen (PSA) PPV: 30%

Digital rectal exam PPV: 28%

Randomized controlled trials Surgery vs. watchful waiting: 2 studies▪ RP effective for clinically detected cancers▪ Only for men < 65

▪ RP not effective for screen-detected cancers▪ Possible benefit for high-risk cancers

Radiation vs. watchful waiting: 1 study▪ No benefit

Dahabrah IJ. Ann Intern Med 2012;156:582

High burden of disease

Long asymptomatic stage

Available screening tests

Available treatments

European Randomized Study of Screening for Prostate Cancer (ERSPC)

Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO)

Randomized controlled trial of 182,160 men ages 50-74 7 European countries

Screening group PSA every 4 years

Control group Usual care

Schröder FH. N Engl J Med 2009; 360:1320

Biopsy referral PSA > 3 ng/mL

Treatment Local standards

Endpoints (9-year follow up) Cancer incidence and mortality

Schröder FH. N Engl J Med 2009; 360:1320

Initial report 162,243 in core age group 55-69

Prostate cancer incidence Screening: 8.2% Control: 4.8% ▪ Rate ratio = 1.70 (95% CI, 1.64 – 1.77)

Schröder FH. N Engl J Med 2009; 360:1320

Relative risk = 0.80 (95% CI 0.65 – 0.98)

Relative risk reduction: 20%Absolute risk reduction: 0.71

deaths/1000 menNeed to invite 1410 men to be

screened twice over 9 years to prevent 1 PCa death Need to diagnose 48 cancers to

prevent 1 PCa death

Schröder FH. N Engl J Med 2009; 360:1320

Randomization methods, screening protocols, and biopsy criteria varied across countries and over time Considered a meta-analysis▪ Positive results only in Netherlands,

Sweden

Unable to exclude treatment effect Control subjects with localized PCa,

particularly with high-risk features, were less likely than screening subjects to receive radical prostatectomy—which is effective.

Control subjects more likely to receive androgen deprivation therapy—which is harmful.

Wolters T. Int J Cancer 2010; 126:2387; Barry MJ. NEJM 2009;360:1351Haines IE. J Natl Cancer Inst 2013;105:1534

Randomized controlled trial of 76,685 men ages 55-74

Screening group Annual PSA and DRE

Control group Usual care

Andriole GL. N Engl J Med 2009; 360:1310

Biopsy referral PSA > 4 ng/mL, abnormal DRE

Treatment Local standards

Endpoints (7-year follow up) Cancer incidence and mortality

Andriole GL. N Engl J Med 2009; 360:1310

Rate ratio = 1.22 (95% CI, 1.16 - 1.29)

Rate ratio = 1.13 (95% CI, 0.75 - 1.70)

Prevalent screening 44% 1+ PSA tests within past 3

yearsContamination in control group

52% underwent PSA testing36% of those with elevated

baseline PSA not biopsied within 3 years (Pinsky PF. J Urol 2005;173:746)

American Cancer Society (ACS)

American College of Physicians (ACP)

American Urological Association (AUA)

United States Preventive Services Task Force (USPTF)

Shared/informed decision making Address screening average-risk men at

50/55▪ 10- to 15-year life expectancy

DRE optional Consider 2-year screening interval

Wolf AMD. Ca Cancer J Clin 2010;60:70; Qaseem A. Ann Intern Med 2013;158:761; Carter HB. J Urol 2013;190:419

Grade D recommendation The USPSTF recommends against

providing [PSA screening] to men without suspicious symptoms regardless of age, race, or family history

An individual man may choose to be screened. The decision should be an informed decision, preferably made in consultation with a regular care provider.

Moyer VM. Ann Intern Med 2012;157:120

Don’t ask, don’t tell (unless they ask)

Complex decisions Multiple acceptable options

involving significant tradeoffs among different possible outcomes

Extensive effect on the patient Controversial

Braddock CH. JAMA 1999; 282:2313

Elements required for complex decisions Discuss▪ Patient’s role in decision making ▪ Clinical issue or nature of decision▪ Alternatives ▪ Potential benefits and harms of the

alternatives▪ Uncertainties associated with the

decisionBraddock CH. JAMA 1999; 282:2313

Collaborative process that allows patients and their providers to make health care decisions together, taking into account the best scientific evidence available, as well as the patient’s values and preferences

http://informedmedicaldecisions.org/

PSA screening is controversial

For most men, chances of harms of screening outweigh chances benefits

Most prostate cancers are indolent

Most men, even if not screened, will not be diagnosed with or die from prostate cancer

Qaseem A. Ann Intern Med 2013;158:761

PSA testing increases risk of cancer diagnosis

PSA does not identify high-risk cancers

Small potential benefit

Many potential harms

Not “just a blood test”Qaseem A. Ann Intern Med 2013;158:761

Benefits (screening every 1 to 4 y for 10 y)

Men, n

10-year PCa death no screening 5 in 1000

10-year PCa death with screening 4-5 in 1000

Net benefit 0-1 in 1000

Harms (screening every 1 to 4 y for 10 y)

Men, n

False positive test 100-120 in 1000

Prostate cancer diagnosis 110 in 1000

Death (treatment) < 1 in 1000

Urinary incontinence (treatment) 18 in 1000

Erectile dysfunction (treatment) 29 in 1000Moyer VA. Ann Intern Med 2012;157:120

Shared decision making Given the complexity of the issues

involved and the time constraints faced by health care providers, we encourage providers and patients to use prostate cancer screening decision aids to facilitate the process

Wolf AMD. Ca Cancer J Clin 2010;60:70

Tools to support decision making when evidence is uncertain and/or very sensitive to patient preferences

Formats: written, video, interactive computer programs, Web-based

Rimer BK. Cancer 2004; 101:1214. Barry MJ. Ann Intern Med 2002; 136:127

Provide evidence-based information about a health condition, the options, associated benefits, harms, probabilities, and uncertainties

O ’Connor AM. Cochrane Database Syst Rev 2009;Jul 8;(3):CD001431

Help patients to recognize the values-sensitive nature of the decision and clarify their preferences

O ’Connor AM. Cochrane Database Syst Rev 2009;Jul 8;(3):CD001431

Provide structured guidance in the steps of decision making and communicating informed values

O ’Connor AM. Cochrane Database Syst Rev 2009;Jul 8;(3):CD001431

Systematic review (18 randomized trials) Increase knowledge Reduce decisional conflict Decrease testing interest ▪ Relative risk = 0.88 (95% CI, 0.81 - 0.97)

Volk RJ. Am J Prev Med 2007;33:428

USPSTFhttp://www.uspreventiveservicestaskforce.org

American College of Physicians Guidance Statementshttp://www.acponline.org/clinical_information/guidelines/guidance/

Independent panel of nonfederal experts in prevention and evidence-based medicine Volunteer members represent

primary care disciplines▪ No substantial financial, intellectual, or

other conflicts that would impair the scientific integrity of the work of the Task Force

Rigorous review of existing peer-reviewed evidence Ratings reflect the strength of the

evidence on the harms and benefits of a preventive service▪ Task Force does not consider the costs

of providing the service or make recommendations for coverage

Grade

Evidence Recommendation

A High certainty of substantial net benefit Provide

B High certainty of moderate net benefitModerate certainty of moderate/substantial net benefit

Provide

C Moderate certainty that net benefit is small

Selectively offer/provide

D No benefit or harms outweigh benefits Do not provide

I Insufficient evidence regarding balance of benefits and harms

Patient Protection and Affordable Care Act Requires private insurers and Medicaid

to cover preventive services that have a grade of “A” or “B” from the USPSTF

Secretary of HHS can modify Medicare coverage of prevention services to be consistent with USPSTF

Grade A Cervical cancer Colorectal cancer

Grade B Breast cancer Lung cancer (LDCT)

Grade I Bladder cancer Skin cancer Oral cancer

Grade D Ovarian cancer Pancreatic cancer Prostate cancer Testicular cancer

Screening programs have important clinical and public health implications

Screening programs should be based on Burden of suffering Natural history Diagnostic tests Treatments

Screening efficacy is best demonstrated by randomized controlled trials showing Decreased mortality Decreased morbidity Preventing disease (sometimes)

Absolute benefits of screening are small

Most people face only potential harms from screening

Physicians should support patients in making informed decisions about cancer screening

Screening guidelines change Use USPSTF, ACP to prepare for

Boards