PPI Trivia What year were PPIs introduced in the U.S.? First PPI - - PDF document

9/29/2016 1

Should we heed the renal failure warnings associated with proton pump inhibitors (PPIs)?

Mary Vilay, PharmD (mvilay@salud.unm.edu) NMSHP Balloon Fiesta Symposium October 3, 2016

PPI Trivia

• What year were PPIs introduced in the U.S.?
• First PPI marketed in the U.S.?
• What year did PPIs become available OTC?

9/29/2016 2

76 year old female

• HTN – treated with

amiloride and HCTZ for several years

• Reflux esophagitis x 1 yr

– Initially treated with famotidine 20 mg daily, but developed recurrent esophageal ulceration with stricture – 6 mo ago, Rx omeprazole 20 mg daily, which was increased to 40 mg daily – Responded well, sx‐free x 3 mo, omeprazole dose lowered to 20 mg daily

• Experiencing increasing

generalized malaise, fatigue, anorexia x 2 wk

• Patient d/c amiloride and

HCTZ x 5 days

• Presents to hospital
• BP 120/60, HR 70 (lying)
• BP 84/56; HR 110 (standing)
• Good skin turgor
• Moist mucous membranes
• No skin rashes
• No flank tenderness
• Ruffenack. Am J Med 1992;93:472.

76 year old female

• Labs:

– Hct 33.2% – WBC 10.3x103/uL – Na 136 mmol/L – K 4.7 mmol/L – Cl 103 mmol/L – CO2 17 mmol/L – BUN 84 mg/dL (19 mg/dL) – SCr 7.2 mg/dL (1.2 mg/dL) – Ca 8.3 mg/dL – Phosphate 6.1 mg/dL – Albumin 2.9 g/dL

• Urinalysis:

– pH 6.5 – Specific gravity 1.006 – Trace protein – 35 WBCs/hpf – No renal tubular epithelial cells – No red blood cells – Wright’s stain – 6% eosinophils

• Renal Ultrasound

– No hydronephrosis

• Ruffenack. Am J Med 1992;93:472.

9/29/2016 3

PPI associated with a number of adverse effects

Could pills for heartburn give you kidney problems? Proton pump inhibitors (PPIs) may come with worrying health effects

9/29/2016 4

Presentation Objectives

Pharmacists 1. Describe how PPIs cause acute kidney injury. 2. Evaluate recent studies demonstrating association between PPIs & renal injury. 3. Formulate recommendations about PPI given the recent evidence. Technicians 1. List renal effects associated with PPIs. 2. State limitations of recent studies investigating renal effects of PPIs. 3. Specify alternative medications for PPIs that are not associated with kidney injury.

PROTON PUMP INHIBITOR USE AND THE RISK OF CHRONIC KIDNEY DISEASE

Lazarus B, et al. JAMA Intern Med 2016 Feb; 176(2): 238‐46

9/29/2016 5

Lazarus et al Study Design

• Objective: quantify association between PPI use

and incident kidney disease in general population

• Secondary outcome: evaluate association

between PPI use and AKI

• Observational cohort study
• Data sources

– Atherosclerosis in Risk in Communities (ARIC) study – Geisinger Health System

Aric Population‐based Cohort

• 15,792 adults

– 45 to 64 years old – Prospectively recruited

• Forsyth, NC
• Jackson, MS
• Suburban Minneapolis, MN
• Washington County, MD
• Participants monitored

via:

– Annual telephone survey – Reviewed community hospital discharge lists – Until Dec. 31, 2011

• Deaths identified by:

– Telephone survey of alternative contacts – Surveillance of public records: newspaper

• bituaries, state death

lists, & death certificates from Dept of Vital Statistics

9/29/2016 6

Participants for PPI Study = ARIC Study Visit 4 Participants

• Visit 4 conducted Feb. 1, 1996 to Jan 30,1999 (N=11,656)
• CKD Analysis

– ACR first obtained at this visit & few patients took PPI before 1996 – Exclusion: missing eGFR or UACR (n=215); eGFR <60 (CKD‐EPI) (n=725)

• Missing data for education, health insurance status, cigarette smoking,

BMI, mean resting SBP, use of antihypertensive or anticoagulant medications, prevalent hypertension, DM, cardiovascular disease (n=234)

– N=10,482

• Dates of study analysis: Feb 1, 1996 to Dec 31, 2011

– Median follow up = 13.9 y

Participants for PPI Study = ARIC Study Visit 4 Participants

• Visit 4 conducted Feb. 1, 1996 to Jan 30,1999
• Visit 4: N=11,656
• AKI analysis: excluded persons with known

ESRD or eGFR <15

– N=11,145

9/29/2016 7

How were outcomes determined?

• Incident CKD defined by diagnostic codes that

indicated CKD at hospital discharge (ICD‐9) or death (ICD‐10) or by incident ESRD (linkage with USRDS)

• Incident AKI defined by hospitalization or death

– ICD‐9 & ICD‐10 codes for acute kidney/renal failure

• Participants who died before developing CKD,

were lost‐to follow‐up or had disease‐free survival were censored

Measurement of PPI in ARIC

• PPIs and H2RAs measured at baseline visit (Jan

1987 to March 1990) via direct visual inspection

• f pill bottles for all medication used in the

preceding 2 weeks

• Subsequent PPI & H2RA exposure obtained

during annual telephone follow‐up

• 2006 onward, participants asked to assemble all

medications and to “read names of all medications prescribed by a doctor”

• Exposure to other medications similarly

measured

9/29/2016 8

Replication Cohort

• Geisinger Health System = large rural health

care system in central & NE Pennsylvania

• Receiving care Feb 13, 1997 to Oct 9, 2014
• Out‐patient eGFR ≥60
• Selection based on earliest time point with

both SCr and SBP available

• N=248,751

– Median follow up 6.2 years

Geisinger Analysis

• Incident CKD = 1st outpatient eGFR <60 sustained at

subsequent eGFR assessments or development ESRD (linkage to USRDS)

• Incident AKI = ICD‐9 code and death (linkage to

National Death Index)

• Individuals who did not develop outcomes censored at

last follow‐up or death

• Medication use determined by prescriber Rx within 90

days before baseline

• PPI frequency categorized as daily or BID (assumed to

be daily if not specified)

• Comorbidities captured by billing codes (in‐pt & out‐pt)

9/29/2016 9

ARIC Baseline Characteristics

Variable PPI Users H2RA Users Non‐Users P‐value Age 62.8±5.5 63.1±5.5 62.5±5.6 0.008 Male 42.5% 39.3% 44.4% 0.01 White 86% 84.2% 77.9% <0.001 Education ≥12 y 81.7% 79.4% 81.8% NS Health Insurance 92.2% 88.9% 85.6% <0.001 Mean eGFR 87.8±13.4 86.5±13.5 88.9±13.1 <0.001 UACR 4 (2‐7.5) 3.6 (1.8‐7.1) 3.7 (1.7‐7.5) NS Cigarette Smoker Current 11.5% 15.5% 15.2% Former 48.4% 44.2% 43.2% NS Never 40.1% 40.3% 41.6%

ARIC Baseline Characteristics

Variable PPI Users H2RA Users Non‐Users P‐value Mean BMI 29.4 (5.3) 29.4 (5.8) 28.7( (5.6) <0.001 SBP 126.5±18.3 128.2±18.6 127±18.8 NS HTN 54.3% 50% 44.8% <0.001 DM 14.9% 18% 15.6% NS CVD 13.7% 14.1% 10.9% 0.003 Medications Antihypertensive 55.3% 48.5% 39.9% <0.001 ACE‐I/ARB 16.8% 13.4% 12.9% NS Diuretic 16.1% 12.1% 9.6% <0.001 Aspirin 64.9% 67.6% 54.9% <0.001 NSAID 27.6% 32.8% 33.2% NS Statin 20.2% 13.6% 10.3% <0.001 Anticoagulant 1.9% 2.8% 1.7% 0.04

9/29/2016 10

Geisinger Cohort Baseline Characteristics

Variable PPI Users H2RA Users Non‐Users P‐value Age 50.0±15.9 50.3±16.3 49.5±16.3 <0.001 Male 43.2% 42.6% 43.5% NS White 94.6% 96.4% 95.5% <0.001 Education ≥12 y NA NA NA NA Health Insurance NA NA NA NA Mean eGFR 94.9±17.7 95.2±18.2 96±18 NA UACR NA NA NA NA Cigarette Smoker Current 25.7% 26.1% 23.9% Former 26.4% 25.4% 23.9% <0.001 Never 47.9% 48.5% 52.2%

Geisinger Cohort Baseline Characteristics

Variable PPI Users H2RA Users Non‐Users P‐value Mean BMI 30.8±7.3 30.8±7.4 30.2±7.1 <0.001 SBP 126.4±15.8 128.2±16.7 128±17.7 <0.001 HTN 33.3% 34% 30.2% <0.001 DM 10.8% 9.7% 10.4% NS CVD 11.3% 11.8% 8.7% <0.001 Medications Antihypertensive 32% 31.3% 20.6% <0.001 ACE‐I/ARB 15.5% 13.4% 9.6% <0.001 Diuretic 13.8% 12.6% 8.3% <0.001 Aspirin 7.8% 5.9% 3.9% <0.001 NSAID 13.9% 14.4% 9.5% <0.001 Statin 13.9% 11.7% 6.1% <0.001 Anticoagulant 2.5% 2.9% 1.1% <0.001

9/29/2016 11

Prevalence PPI Ever Use Over Time in ARIC Study

PPI Use & Risk of Incident CKD

ARIC (N=10,482) Geisinger (N=248,751) HR (95% CI) HR (95% CI) Unadjusted baseline PPI use vs non‐users 1.45 (1.11 – 1.90) 1.20 (1.15 – 1.26) Adjusted baseline PPI use vs non‐users 1.50 (1.14 – 1.96) 1.17 (1.12 – 1.23) Time‐varying PPI ever use vs non‐users 1.35 (1.17 – 1.55) 1.22 (1.19 – 1.25) Baseline PPI use vs baseline H2RA use 1.39 (1.01 – 1.91) 1.29 (1.19 – 1.14) Baseline PPI use vs propensity score‐ matched non‐users 1.76 (1.13 – 2.74) 1.16 (1.09 – 1.24) Time varying PPI ever use vs never PPI use, excluding baseline PPI users NA 1.24 (1.20 – 1.28) Baseline PPI use vs non‐users (excluding patients with albuminuria) 1.45 (1.09‐1.96) 1.19 (1.113‐1.25) Negative Control Baseline H2RA use vs non‐user 1.15 (0.98 – 1.36) 0.93 (0.88 – 0.99)

9/29/2016 12

Adjustment Variables

ARIC

• Age
• Sex
• Race
• Study center
• Education
• Health insurance status
• Baseline eGFR
• UACR
• Smoking status
• BMI
• SBP
• DM
• Cardiovascular disease
• Antihypertensive medication use
• Anticoagulant medication use

Geisinger Health

• Age
• Sex
• Race
• Baseline eGFR
• Smoking status
• BMI
• SBP
• DM
• Cardiovascular disease
• Antihypertensive medication use
• Anticoagulant medication use
• Statin use
• ASA use
• NSAID use
• Propensity score‐matched analyses were

adjusted for propensity scores only, estimated using same variables

PPI Use & Risk of Incident AKI

ARIC (N=10,482) Geisinger (N=248,751) HR (95% CI) HR (95% CI) Unadjusted baseline PPI use vs non‐users 1.72 (1.28 – 2.30) 1.30 (1.21 – 1.40) Adjusted baseline PPI use vs non‐users 1.64 (1.22 – 2.21) 1.31 (1.22 – 1.42) Time‐varying PPI ever use vs non‐users 1.49 (1.25 – 1.77) 1.54 (1.47 – 1.60) Baseline PPI use vs baseline H2RA use 1.58 (1.05 – 2.40) 1.30 (1.13 – 1.48) Baseline PPI use vs propensity score‐ matched non‐users 2.00 (1.24 – 3.22) 1.29 (1.16 – 1.43) Time varying PPI ever use vs never PPI use, excluding baseline PPI users NA 1.66 (1.57 – 1.75) Negative Control Baseline H2RA use vs non‐user 1.03 (0.84 – 1.26) 0.98 (0.89 – 1.10)

9/29/2016 13

Adjusted HR based on Dose Frequency (Geisinger Health System Data)

1 2 3

CKD Daily PPI BID PPI AKI Daily PPI BID PPI

Study Limitations

• Risk of unaccounted factors
• Surveillance bias
• Low sensitivity of hospital discharge codes for

diagnosing CKD in ARIC study

• Inclusion of baseline PPI users (selecting those

tolerant to PPI) can invoke selection bias

• PPI/H2RA use not captured by directly
• bserved therapy

– Exposure misclassification

9/29/2016 14

Strengths

• ARIC – large representative community‐based sample

– Baseline (visit 4) occurring soon after PPIs introduced in U.S. – Visual confirmation of medications used – Comprehensive data pertaining to potential confounders – Close monitoring with >13 y follow‐up

• Sensitivity analyses, including time‐varying exposure

model, propensity score matching

• Replicated results in a large second cohort (Geisinger)
• Demonstrated specificity to PPI rather than H2RA use

Lazarus et al. JAMA Intern Med 2016 Summary

• PPI use is associated with incident CKD and

AKI in the general population

– Association with incident CKD appears to be dose dependent

9/29/2016 15

ASSOCIATION BETWEEN USE OF PPI AND RISK OF ESRD IN RENAL DISEASES

Peng YC et al. Medicine 2016 Apr; 95(15): e3363

Taiwanese Study

• Objective: analyze risk of developing ESRD

among patients with renal disease who use PPIs

• Case‐controlled study
• Data source: Taiwan National Health Insurance

Research Database (NHIRD)

– Compulsory, social insurance program – Provides healthcare for >99% of residents since 1995

9/29/2016 16

Subject Selection

• ICD‐9 Codes
• Focused on patients with renal disease:

– Nephritis, nephritic syndrome, gomerulonephritis, nephropathy, CKD, renal function impairment

• Matched controls 1:1 (propensity score)
• PPI prescriptions require diagnosis of PUD,

reflux esophagitis, GERD (diagnosed by upper GI endoscopy)

2006

9/29/2016 17

Baseline Characteristics

Control N=3808 ESRD N=3808 P‐value Gender NS Women 47.9% 47.8% Men 52.2% 52.2% Mean Age 66.1±13. 8 65.4±13. 1 0.03 Co‐morbidities DM 46.5% 46.5% NS HTN 83.6% 81.7% 0.02 COPD 44.4% 42.8% NS CAD 50.7% 49.1% NS CHF 22.0% 21.2% NS Cancer 9.7% 9.3% NS Control N=3808 ESRD N=3808 P‐value Medications PPI 55.2% 69.5% <0.001 H2RA 87.8% 86.5% NS ACE‐I 49.6% 49.1% NS ARB 40.8% 41.4% NS Beta blocker 77.0% 76.3% NS CCB 89.2% 86.8% NS Aldoster ‐one inhibitor 24.8% 24.1% NS NSAID 98.1% 98.3% NS

ESRD with PPI Use

Case #/ Control # Adjusted OR* 95% 1.36CI PPI 2647/2104 1.88 (1.71, 2.06) Omeprazole 986/900 1.14 (1.02, 1.26) Pantoprazole 443/286 1.63 (1.39, 1.90) Lansoprazole 654/551 1.23 (1.09, 1.40) Rabeprazole 112/83 1.36 (1.02, 1.81) Esomeprazole 452/282 1.69 (1.45, 1.98) * Adjusted for gender, age, CCB, DM, HTN

9/29/2016 18

ESRD Associated with Cumulative Defined Daily Dose

Case #/Control # Adjusted OR 95% CI PPI <100 DDD 2022/1562 1.92 (1.74, 2.13) PPI ≥100 DDD 625/540 1.74 (1.52, 2.00) Omeprazole <80 DDD 792/679 1.60 (1.41, 1.82) Omeprazole ≥80 DDD 1916/1423 2.01 (1.81, 2.22) Pantoprazole <70 DDD 350/202 2.58 (2.14, 3.12) Pantoprazole ≥ 70 DDD 2297/1900 1.80 (1.64, 1.99) Lansoprazole <110 DDD 449/397 1.87 (1.61, 2.18) Lansoprazole ≥110 DDD 2148/1705 1.88 (1.70, 2.07) Rabeprazole <105 DDD 38/25 2.19 (1.31, 3.65) Rabeprazole ≥105 DDD 2569/2048 1.87 (1.70, 2.06) Esomeprazole <140 DDD 338/208 2.41 (2.00, 2.91) Esopmeprazole ≥140 DDD 2309/1894 1.82 (1.65, 2.00)

Limitations

• CKD Stage and SCr not available through NHIRD
• Variable ranges with PPI dose
• Assumed patients ingested PPI as written on Rx
• OTC PPI may have impacted results, OTC

information not captured in NHIRD

• NHIRD not completely comprehensive: missing

information on factors that affect renal function e.g. BP, proteinuria, HgA1C, uric acid, iPTH, CBC

• Unknown time between renal disease to ESRD

with PPI use

9/29/2016 19

Peng et al. Medicine 2016 Summary

• PPI associated with risk of developing ESRD

inpatients with renal disease

• Risk of developing ESRD associated with each

PPI individually

• Risk of developing ESRD may be dose related

J Am Soc Nephrol 2016 Apr; [Epub ahead of print]

9/29/2016 20

Data Sources

• VA databases (in‐pt & out‐pt encounters)

– Pt demographics & comorbidity using Current Procedural Terminology codes & ICD‐9 codes

• VA Managerial Cost Accounting System Laboratory Results

– Out‐pt & in‐pt lab results

• VA corporate Data Warehouse Production Outpatient

Pharmacy

– Rx info

• VA Vital Status & Beneficiary ID Records Locator System

– Demographic characteristics & death through Sept 30, 2013

• USRDS

– ESRD occurrence & date of first ESRD services

VA Users after Oct 1, 2006 N=8,434,579 No PPI Rx Oct 1, 1999 – Sept 30, 2006 N=7,188,415 PPI Rx Oct 1, 2006 – Sept 30,2008 N=371,496 Baseline eGFR within 90 days pre 1st PPI Rx & another eGFR after 1st PPI RX N=212,181 Baseline eGFR ≥60 N=173,321 No PPI Rx Oct 1 1999 – Sept 30, 2013 N=5,918,697 No H2 Rx before Sept 30, 2006 N=5,707,766 H2 Rx Oct 1,2006 – Sept 30, 2008 N=45,514 Baseline eGFR within 90 days pre 1st H2 Rx & another eGFR after 1st H2 RX N=24,145 Baseline eGFR ≥60 N=20,270

9/29/2016 21

Outcome Definitions

• Primary outcome = eGFR <60 mL/min/1.73 m2 (CKD‐EPI)
• CKD = TWO eGFRs <60, ≥90 days apart
• CKD occurrence = first date eGFR <60
• CKD Progression:

– >30% decline in eGFR – Doubling SCr – ESRD – ESRD or >50% decline eGFR – Outcomes based on outpatient SCr, except ESRD

• Outcomes ascertained for 5 years from time of cohort entry
• r death

Baseline Demographics

Baseline Characteristics H2RA N=20,270 PPI N=173,321 P‐value Age 55.40±12.81 56.85±11.85 <0.001 Baseline eGFR 89.98±15.88 86.56±15.67 <0.001 Race 0.01 White 78.62% 79.14% Black 18.67% 18.47% Other 2.71% 2.38% Sex NS Male 93.38% 93.04% Female 6.62% 6.96% Years of follow‐up 5.00 (IQR 5.00, 5.00) 5.00 (IQR 5.00, 5.00) <0.001

9/29/2016 22

Baseline Medical Conditions

Baseline Characteristics H2RA N=20,270 PPI N=173,321 P‐value DM 44.02% 41.72% <0.001 HTN 78.02% 78.2% <0.01 Chronic Lung Dz 39.23% 38.63% NS PAD 24.71% 18.07% <0.001 Cardiovascular Dz 41.73% 41.43% NS Cerebrovascular Dz 22.67% 15.26% <0.001 Dementia 24.95% 18.68% <0.001 Hyperlipidemia 72.94% 73.54% NS HCV 5.91% 8.59% <0.001 HIV 0.27% 0.39% <0.01

Baseline GI Conditions

Baseline Characteristics H2RA N=20,270 PPI N=173,321 P‐value GERD 18.58% 50.08% <0.001 UGIB 1.21% 4.56% <0.001 PUD 3.29% 15.13% <0.001

• H. Pylori

0.11% 2.34% <0.001 Barrett esophagus 0.07% 1.85% <0.001 Achlasia 0.00% 0.12% <0.001 Stricture 0.16% 1.33% <0.001 Esophageal adenocarcinoma 0.01% 0.17% <0.001 Days of having related Rx during follow‐up 90 (IQR 30, 270) 450 (IQR 90, 1260) <0.001

9/29/2016 23

Association PPI and Renal Outcomes

Outcome H2RA (N=20,270) PPI (N=173,321) Incident eGFR <60 mL/min/1.73 m2 # of events (%) 4,429 (21.85%) 48,171 (27.79) Incident rate per 100,000 person years (95% CI) 5408.24 (5248.96 to 5567.52) 7241.27 (7176.61 to 7305.94) Adjusted HR (95% CI) 1.0 1.22 (1.18 to 1.26) Incident CKD # of events (%) 2,234 (11.02) 26,193 (15.11) Incident rate per 100,000 person years (95% CI) 2569.86 (2463.3 to 2676.43) 3683.12 (3638.52 to 3727.72) Adjusted HR (95% CI) 1.0 1.28 (1.23 to 1.34)

Covariates used to adjust models

• Baseline eGFR
• Age
• Race
• Sex
• DM
• HTN
• Cardiovascular Dz
• PAD
• Cerebrovascular Dz
• Chronic Lung disease
• HCV
• HIV
• Dementia
• Reason for PPI use:

– GERD – UGIB – PUD – H. pylori – Barett – Achlasia stricture – Esophageal adenocarcinoma

9/29/2016 24

Association PPI and CKD Progression

Outcome H2RA (N=20,270) PPI (N=173,321) HR (95% CI) Adjusted Doubling SCr 1.0 1.53 (1.42 to 1.65) >30% decline eGFR 1.0 1.32 (1.28 to 1.37) ESRD 1.0 1.96 (1.21 to 3.18) ESRD or >50% decline eGFR 1.0 1.47 (1.38 to 1.57)

Duration PPI Exposure & Risk of Renal Outcomes (N=173,321)

9/29/2016 25

1:1 propensity score‐matched cohort

• f PPI new‐users vs H2RA new‐users

Outcome PPI (N=20,270) vs H2 (N=20,270) HR (95% CI) Incident eGFR <60 mL/min/1.73 m2 1.23 (1.17 to 1.30) Incident CKD 1.28 (1.18 to 1.38) Doubling of SCr 1.63 (1.47 to 1.81) >30% decline in eGFR 1.32 (1.25 to 1.39) ESRD 1.48 (0.49 to 4.50) ESRD or >50% decline in eGFR 1.59 (1.45 to 1.74)

1:1 propensity score‐matched cohort

• f PPI new‐users vs Control

Outcome PPI (N=173,321) vs Control (N=173,321) HR (95% CI) Incident eGFR <60 mL/min/1.73 m2 1.57 (1.54 to 1.60) Incident CKD 1.81 (1.76 to 1.86) Doubling of SCr 1.86 (1.80 to 1.93) >30% decline in eGFR 1.67 (1.64 to 1.70) ESRD 1.61 (1.26 to 2.04) ESRD or >50% decline in eGFR 1.83 (1.77 to 1.89)

9/29/2016 26

PPI Use and Risk AKI

• HR 2.15 (2.00 to 2.32)

Models additionally adjusted for AKI during exposure to acid‐suppression

Outcome PPI (N=173,321) vs H2 (N=20,270) HR (95% CI) Incident eGFR <60 mL/min/1.73 m2 1.20 (1.16 to 1.24) Incident CKD 1.28 (1.22 to 1.34) Doubling of SCr 1.42 (1.32 to 1.54) >30% decline in eGFR 1.28 (1.24 to 1.33) ESRD 1.79 (1.10 to 2.89) ESRD or >50% decline in eGFR 1.38 (1.29 to 1.47)

9/29/2016 27

Sensitivity Analysis

• Inclusion of additional covariate:

– Number eGFR measurements – NSAD use for ≥30 days before or during study timeframe – Serum sodium bicarbonate – ACEI and ARB use for ≥30 days before or during study timeframe

• Results remained consistent

Limitations

• Cohort focused on older, white, male, U.S.

veterans → limits generalizability

• Sampling bias and inaccurate measurement of

predictor variables

• Did not capture OTC PPI – likely minimal

effect, would have biased against (underestimation) risk with PPI

• **98.6% PPI exposure = omeprazole
• **97.9% H2RA exposure = ranitidine

9/29/2016 28

Strengths

• National large‐scale data from network of

integrated health systems captured during roune medical care → minimizes selecon bias

• Evaluated multiple outcomes in continuum CKD

evolution

– CKD onset – CKD progression – ESRD onset

• Different cohort designs and models to test

sensitivity

Xie et al. JASN 2016 Summary

• PPI associated with increased risk:

– Incident CKD – CKD progression – ESRD

• Graded association of PPI exposure and risk of

renal outcomes

• PPI association with renal diseases may be

independent of AKI development

9/29/2016 29

How to apply results to clinical practice?

• IMS Institute for Healthcare Informatics (2013)

– >15 million Americans use prescription PPIs costing > $10 billion