Analytical Considerations When Monitoring Pain Medications by LC-MS/MSDavid Masters-Moore
Laboratory urine drug testing of patients on chronic opioid therapy requires providing a large test menu of medications commonly prescribed for this population as well as metabolites and illicit substances. It has been shown that liquid chromatography-tandem mass spectrometry (LC-MS/MS) is the preferred method to analyze urine specimens for these substances.
Purpose of the study: To describe the challenges and some of the techniques to validate the analytical procedures used to identify and quantify these medications and substances.
Methods: Using data obtained from testing over one million specimens, the authors developed a proposed test menu. Potential isobaric interferences were established by using literature references. A list of potentially interfering medications was obtained by using the proposed test menu and the most commonly prescribed medications. Finally, criteria were designed to detect possible carryover.
Results: The LC-MS/MS instrumentation eliminated all potential interferences and provided quantitative data over the test range needed to monitor these patients. Carryover could be eliminated by setting the carryover thresholds for each analyte.
Conclusions: Reference laboratories utilizing LC-MS/MS technology to conduct urine drug testing for pain clinicians should employ specific techniques described in this study to develop an optimal test menu and validate procedures that include isolating retention times for isobaric compounds, identifying interfering substances including impurities in medicinal and illicit substance preparations, monitoring ion suppression, and avoiding carryover.
ACTIVE MMP 8 IN DIAGNOSIS OF PERIODONTAL DISEASE PROGRESSION. COLLAGENASE 2 (MMP8) IN PERIODONTAL DISEASE. PERIODONTAL DISEASE BIOMARKER. POC (POINT OF CARE) TESTING. CHAIR SIDE TESTS. HYBRID LABORATORY.
Analytical Considerations When Monitoring Pain Medications by LC-MS/MSDavid Masters-Moore
Laboratory urine drug testing of patients on chronic opioid therapy requires providing a large test menu of medications commonly prescribed for this population as well as metabolites and illicit substances. It has been shown that liquid chromatography-tandem mass spectrometry (LC-MS/MS) is the preferred method to analyze urine specimens for these substances.
Purpose of the study: To describe the challenges and some of the techniques to validate the analytical procedures used to identify and quantify these medications and substances.
Methods: Using data obtained from testing over one million specimens, the authors developed a proposed test menu. Potential isobaric interferences were established by using literature references. A list of potentially interfering medications was obtained by using the proposed test menu and the most commonly prescribed medications. Finally, criteria were designed to detect possible carryover.
Results: The LC-MS/MS instrumentation eliminated all potential interferences and provided quantitative data over the test range needed to monitor these patients. Carryover could be eliminated by setting the carryover thresholds for each analyte.
Conclusions: Reference laboratories utilizing LC-MS/MS technology to conduct urine drug testing for pain clinicians should employ specific techniques described in this study to develop an optimal test menu and validate procedures that include isolating retention times for isobaric compounds, identifying interfering substances including impurities in medicinal and illicit substance preparations, monitoring ion suppression, and avoiding carryover.
ACTIVE MMP 8 IN DIAGNOSIS OF PERIODONTAL DISEASE PROGRESSION. COLLAGENASE 2 (MMP8) IN PERIODONTAL DISEASE. PERIODONTAL DISEASE BIOMARKER. POC (POINT OF CARE) TESTING. CHAIR SIDE TESTS. HYBRID LABORATORY.
Bayesian theory in population pharmacokinetics--
1) INTRODUCTION TO BAYESIAN THEORY
2)BAYESIAN PROBABILITY TO DOSING OF DRUGS
3)APPLICATIONS AND USES OF BAYESIAN THEORY IN APPLIED PHARMACOKINETICS:
therapeutic drug monitoring and clinical pharmacokinetics-fifth pharm d notes
SGS Biopharm Day 2016 - Modeling & simulation in Phase 1Ruben Faelens
This slidedeck shows how Model-Informed Drug Development can be applied in a Phase 1 first-in-human trial, informing starting dose, dose escalation as well as stop dose.
Bayesian estimations of strong toxic signals [compatibility mode]Bhaswat Chakraborty
“Signals” of adverse drug reactions are, according to WHO, “reported information on a possible causal relationship between an adverse event and a drug, the relationship being unknown or incompletely documented previously. Usually more than a single report is required to detect a signal, depending on the seriousness of the event and the quality of the information.” Once a signal is detected, one can then analyze and confirm it. In detecting signals from large adverse drug reaction (ADR) databases, however, one has to use a procedure that is sensitive (low false negativity) and specific (high true positivity) for the purpose. A whole range of statistical methods have been applied for data mining and signal detection (SD) in pharmacovigilance (PV). My talk would be on Bayesian methods for SD.
The US FDA uses a Bayesian data mining approach developed by William DuMouchel called multi-item gamma poisson shrinker (MGPS). WHO also uses a Bayesian method (Andrew Bate) based on a Bayesian confidence propagation neural network (BCPNN). These estimates provide shrinkage towards zero of the observed to expected number of ADRs, e.g., the empirical Bayesian geometric mean (EBGM) or information component (IC). These Bayesian estimators are robust measures of ADR-drug association.
Bayesian approaches are intuitively appealing when very small numbers are involved and where there is a need of continuous reassessment of probability of association with the acquisition of new data over time. Bayesian estimates such as EBGM are close to null hypothesis of independence even when the data is scarce. For example, if the EBGM is 5 for a drug-renal toxicity combination, then this drug-event combination occurred, on an average, 5 times more frequently than expected in the data set. Several examples Bayesian SD will be given from current research projects.
Classifying Readmissions of Diabetic Patient EncountersMayur Srinivasan
Readmission rates in hospitals are a key indicator on quality of patient care and a clear indication of total cost or inconvenience related to the treatment. Patients with serious medical
conditions such as diabetes mellitus are key drivers of readmission rates owing to the complexity of their illness. Therefore, being able to predict based on certain features whether or not a patient
will need readmission can help doctors and hospitals provide better care initially and not get financially penalized under Obamacare’s readmission policy
Bayesian theory in population pharmacokinetics--
1) INTRODUCTION TO BAYESIAN THEORY
2)BAYESIAN PROBABILITY TO DOSING OF DRUGS
3)APPLICATIONS AND USES OF BAYESIAN THEORY IN APPLIED PHARMACOKINETICS:
therapeutic drug monitoring and clinical pharmacokinetics-fifth pharm d notes
SGS Biopharm Day 2016 - Modeling & simulation in Phase 1Ruben Faelens
This slidedeck shows how Model-Informed Drug Development can be applied in a Phase 1 first-in-human trial, informing starting dose, dose escalation as well as stop dose.
Bayesian estimations of strong toxic signals [compatibility mode]Bhaswat Chakraborty
“Signals” of adverse drug reactions are, according to WHO, “reported information on a possible causal relationship between an adverse event and a drug, the relationship being unknown or incompletely documented previously. Usually more than a single report is required to detect a signal, depending on the seriousness of the event and the quality of the information.” Once a signal is detected, one can then analyze and confirm it. In detecting signals from large adverse drug reaction (ADR) databases, however, one has to use a procedure that is sensitive (low false negativity) and specific (high true positivity) for the purpose. A whole range of statistical methods have been applied for data mining and signal detection (SD) in pharmacovigilance (PV). My talk would be on Bayesian methods for SD.
The US FDA uses a Bayesian data mining approach developed by William DuMouchel called multi-item gamma poisson shrinker (MGPS). WHO also uses a Bayesian method (Andrew Bate) based on a Bayesian confidence propagation neural network (BCPNN). These estimates provide shrinkage towards zero of the observed to expected number of ADRs, e.g., the empirical Bayesian geometric mean (EBGM) or information component (IC). These Bayesian estimators are robust measures of ADR-drug association.
Bayesian approaches are intuitively appealing when very small numbers are involved and where there is a need of continuous reassessment of probability of association with the acquisition of new data over time. Bayesian estimates such as EBGM are close to null hypothesis of independence even when the data is scarce. For example, if the EBGM is 5 for a drug-renal toxicity combination, then this drug-event combination occurred, on an average, 5 times more frequently than expected in the data set. Several examples Bayesian SD will be given from current research projects.
Classifying Readmissions of Diabetic Patient EncountersMayur Srinivasan
Readmission rates in hospitals are a key indicator on quality of patient care and a clear indication of total cost or inconvenience related to the treatment. Patients with serious medical
conditions such as diabetes mellitus are key drivers of readmission rates owing to the complexity of their illness. Therefore, being able to predict based on certain features whether or not a patient
will need readmission can help doctors and hospitals provide better care initially and not get financially penalized under Obamacare’s readmission policy
Pyxl Webinar: Fast Forward: Mobile, Social Media & Video Predictions for 2016Pyxl
Hear Chad Elmore, Pyxl’s director of digital communication, break down mobile, social media and video trends, discuss what you can expect from each in the coming year and give you ideas about how to use each to your advantage.
Precarity Pilot: exceedig precarising models of design practiceBrave New Alps
These slides were part of our presentation at the Annual American Geographers meeting: Making Other Worlds Possible V - The Role of Disruptive Innovation and New Political Imaginaries, Chicago, 21 April 2015
This is a report to Exxon Mobil's shareholders detailing the companies risk management strategy concerning climate change and its oil and gas activities.
Read more: http://on.mash.to/1fOH1xL
Marketing is Dead. Only Moments Matter - UserTesting Roadshow - 10/5/2016Kyle Lacy
Another adjustment to my Marketing is Dead deck which covers how to evolve in the digital environment. The only thing that matters or should matter to digital marketing is the experience the consumer is having with your brand.
Are you interested in learning how to prevent hospital readmissions for your diabetic population? It is a popular belief that measuring blood glucose for your diabetic population is the most predictive variable in determining a hospital readmission for a diabetic. However, many providers of care simply do not perform the test on known diabetic patients. This study takes a look at an advanced analytic method that works within the current healthcare providers workflow to looks to identify the likelihood of a future 30-day unplanned readmission before hospital discharge.
This presentation was made at the PAMM winter meeting in Verona (Italy) February 2019 and intended students to go through the basic methods used for phase I clinical trials.
UPDATED-Early Phase Drug Developmetn and Population PK and Its' ValueE. Dennis Bashaw
Presentation Given at Regional AAPS DDDI Meeting in Baltimore. Similar to previous talks BUT updated to include a discussion of BIA 10-2474 and extended discussion of risk
EHR-based Phenome Wide Association Study in Pancreatic CancerTomasz Adamusiak
Presented at 2014 AMIA Joint Summits, April 9, 2014, San Francisco, CA
BACKGROUND. Pancreatic cancer is one of the most common causes of cancer-related deaths in the United States, it is difficult to detect early and typically has a very poor prognosis. We present a novel method of large-scale clinical hypothesis generation based on phenome wide association study performed using Electronic Health Records (EHR) in a pancreatic cancer cohort. METHODS. The study population consisted of 1,154 patients diagnosed with malignant neoplasm of pancreas seen at The Froedtert & The Medical College of Wisconsin academic medical center between the years 2004 and 2013. We evaluated death of a patient as the primary clinical outcome and tested its association with the phenome, which consisted of over 2.5 million structured clinical observations extracted out of the EHR including labs, medications, phenotypes, diseases and procedures. The individual observations were encoded in the EHR using 6,617 unique ICD-9, CPT-4, LOINC, and RxNorm codes. We remapped this initial code set into UMLS concepts and then hierarchically expanded to support generalization into the final set of 10,164 clinical concepts, which formed the final phenome. We then tested all possible pairwise associations between any of the original 10,164 concepts and death as the primary outcome. RESULTS. After correcting for multiple testing and folding back (generalizing) child concepts were appropriate, we found 231 concepts to be significantly associated with death in the study population.
CONCLUSIONS. With the abundance of structured EHR data, phenome wide association studies combined with
knowledge engineering can be a viable method of rapid hypothesis generation.
The ppt is a short description about how to ascertain the validity, ie; sensitivity and specificity of a screening test as well as their predictive powers. you can also find the technique to ascertain the best possible screening test through the help of an ROC curve...
MedicReS Conference 2017 Istanbul - Fostering Responsible Conduct of Research...MedicReS
Fostering Responsible Conduct of Research
MedicReSConference
May 5, 2017
Istanbul, Turkey
Adil E. Shamoo, Ph.D., CIP
University of Maryland School of Medicine
MedicReS Conference 2017 Istanbul - Ethical issues of secondary analysis of a...MedicReS
Ethical issues of secondary analysis of archived data
MedicReS Conference
May 4, 2017
Istanbul, Turkey
Adil E. Shamoo, Ph.D., CIP
University of Maryland School of Medicine
MedicReS Conference 2017 Istanbul - Integrity of Authorship in Research Publi...MedicReS
Integrity of Authorship in Research Publications
MedicReSConference
May 4, 2017
Istanbul, Turkey
Adil E. Shamoo, Ph.D., CIP
University of Maryland School of Medicine
MedicReS Winter School 2017 Vienna - Ethics of Cancer Trials - Adil E. ShamooMedicReS
A Comprehensive Introduction to the Ethical Issues at stake in the conduct of Cancer Research
Adil E. Shamoo, Ph.D.
University of Maryland School of Medicine
Knee anatomy and clinical tests 2024.pdfvimalpl1234
This includes all relevant anatomy and clinical tests compiled from standard textbooks, Campbell,netter etc..It is comprehensive and best suited for orthopaedicians and orthopaedic residents.
Muktapishti is a traditional Ayurvedic preparation made from Shoditha Mukta (Purified Pearl), is believed to help regulate thyroid function and reduce symptoms of hyperthyroidism due to its cooling and balancing properties. Clinical evidence on its efficacy remains limited, necessitating further research to validate its therapeutic benefits.
ABDOMINAL TRAUMA in pediatrics part one.drhasanrajab
Abdominal trauma in pediatrics refers to injuries or damage to the abdominal organs in children. It can occur due to various causes such as falls, motor vehicle accidents, sports-related injuries, and physical abuse. Children are more vulnerable to abdominal trauma due to their unique anatomical and physiological characteristics. Signs and symptoms include abdominal pain, tenderness, distension, vomiting, and signs of shock. Diagnosis involves physical examination, imaging studies, and laboratory tests. Management depends on the severity and may involve conservative treatment or surgical intervention. Prevention is crucial in reducing the incidence of abdominal trauma in children.
Title: Sense of Taste
Presenter: Dr. Faiza, Assistant Professor of Physiology
Qualifications:
MBBS (Best Graduate, AIMC Lahore)
FCPS Physiology
ICMT, CHPE, DHPE (STMU)
MPH (GC University, Faisalabad)
MBA (Virtual University of Pakistan)
Learning Objectives:
Describe the structure and function of taste buds.
Describe the relationship between the taste threshold and taste index of common substances.
Explain the chemical basis and signal transduction of taste perception for each type of primary taste sensation.
Recognize different abnormalities of taste perception and their causes.
Key Topics:
Significance of Taste Sensation:
Differentiation between pleasant and harmful food
Influence on behavior
Selection of food based on metabolic needs
Receptors of Taste:
Taste buds on the tongue
Influence of sense of smell, texture of food, and pain stimulation (e.g., by pepper)
Primary and Secondary Taste Sensations:
Primary taste sensations: Sweet, Sour, Salty, Bitter, Umami
Chemical basis and signal transduction mechanisms for each taste
Taste Threshold and Index:
Taste threshold values for Sweet (sucrose), Salty (NaCl), Sour (HCl), and Bitter (Quinine)
Taste index relationship: Inversely proportional to taste threshold
Taste Blindness:
Inability to taste certain substances, particularly thiourea compounds
Example: Phenylthiocarbamide
Structure and Function of Taste Buds:
Composition: Epithelial cells, Sustentacular/Supporting cells, Taste cells, Basal cells
Features: Taste pores, Taste hairs/microvilli, and Taste nerve fibers
Location of Taste Buds:
Found in papillae of the tongue (Fungiform, Circumvallate, Foliate)
Also present on the palate, tonsillar pillars, epiglottis, and proximal esophagus
Mechanism of Taste Stimulation:
Interaction of taste substances with receptors on microvilli
Signal transduction pathways for Umami, Sweet, Bitter, Sour, and Salty tastes
Taste Sensitivity and Adaptation:
Decrease in sensitivity with age
Rapid adaptation of taste sensation
Role of Saliva in Taste:
Dissolution of tastants to reach receptors
Washing away the stimulus
Taste Preferences and Aversions:
Mechanisms behind taste preference and aversion
Influence of receptors and neural pathways
Impact of Sensory Nerve Damage:
Degeneration of taste buds if the sensory nerve fiber is cut
Abnormalities of Taste Detection:
Conditions: Ageusia, Hypogeusia, Dysgeusia (parageusia)
Causes: Nerve damage, neurological disorders, infections, poor oral hygiene, adverse drug effects, deficiencies, aging, tobacco use, altered neurotransmitter levels
Neurotransmitters and Taste Threshold:
Effects of serotonin (5-HT) and norepinephrine (NE) on taste sensitivity
Supertasters:
25% of the population with heightened sensitivity to taste, especially bitterness
Increased number of fungiform papillae
Integrating Ayurveda into Parkinson’s Management: A Holistic ApproachAyurveda ForAll
Explore the benefits of combining Ayurveda with conventional Parkinson's treatments. Learn how a holistic approach can manage symptoms, enhance well-being, and balance body energies. Discover the steps to safely integrate Ayurvedic practices into your Parkinson’s care plan, including expert guidance on diet, herbal remedies, and lifestyle modifications.
Rasamanikya is a excellent preparation in the field of Rasashastra, it is used in various Kushtha Roga, Shwasa, Vicharchika, Bhagandara, Vatarakta, and Phiranga Roga. In this article Preparation& Comparative analytical profile for both Formulationon i.e Rasamanikya prepared by Kushmanda swarasa & Churnodhaka Shodita Haratala. The study aims to provide insights into the comparative efficacy and analytical aspects of these formulations for enhanced therapeutic outcomes.
Basavarajeeyam is an important text for ayurvedic physician belonging to andhra pradehs. It is a popular compendium in various parts of our country as well as in andhra pradesh. The content of the text was presented in sanskrit and telugu language (Bilingual). One of the most famous book in ayurvedic pharmaceutics and therapeutics. This book contains 25 chapters called as prakaranas. Many rasaoushadis were explained, pioneer of dhatu druti, nadi pareeksha, mutra pareeksha etc. Belongs to the period of 15-16 century. New diseases like upadamsha, phiranga rogas are explained.
These lecture slides, by Dr Sidra Arshad, offer a quick overview of the physiological basis of a normal electrocardiogram.
Learning objectives:
1. Define an electrocardiogram (ECG) and electrocardiography
2. Describe how dipoles generated by the heart produce the waveforms of the ECG
3. Describe the components of a normal electrocardiogram of a typical bipolar lead (limb II)
4. Differentiate between intervals and segments
5. Enlist some common indications for obtaining an ECG
6. Describe the flow of current around the heart during the cardiac cycle
7. Discuss the placement and polarity of the leads of electrocardiograph
8. Describe the normal electrocardiograms recorded from the limb leads and explain the physiological basis of the different records that are obtained
9. Define mean electrical vector (axis) of the heart and give the normal range
10. Define the mean QRS vector
11. Describe the axes of leads (hexagonal reference system)
12. Comprehend the vectorial analysis of the normal ECG
13. Determine the mean electrical axis of the ventricular QRS and appreciate the mean axis deviation
14. Explain the concepts of current of injury, J point, and their significance
Study Resources:
1. Chapter 11, Guyton and Hall Textbook of Medical Physiology, 14th edition
2. Chapter 9, Human Physiology - From Cells to Systems, Lauralee Sherwood, 9th edition
3. Chapter 29, Ganong’s Review of Medical Physiology, 26th edition
4. Electrocardiogram, StatPearls - https://www.ncbi.nlm.nih.gov/books/NBK549803/
5. ECG in Medical Practice by ABM Abdullah, 4th edition
6. Chapter 3, Cardiology Explained, https://www.ncbi.nlm.nih.gov/books/NBK2214/
7. ECG Basics, http://www.nataliescasebook.com/tag/e-c-g-basics
- Video recording of this lecture in English language: https://youtu.be/kqbnxVAZs-0
- Video recording of this lecture in Arabic language: https://youtu.be/SINlygW1Mpc
- Link to download the book free: https://nephrotube.blogspot.com/p/nephrotube-nephrology-books.html
- Link to NephroTube website: www.NephroTube.com
- Link to NephroTube social media accounts: https://nephrotube.blogspot.com/p/join-nephrotube-on-social-media.html
1. Why most published studies
are wrong but can be fixed
David Madigan
Columbia University
http://www.omop.org
http://www.ohdsi.org
“The sole cause and root of almost every defect in the sciences is this: that whilst
we falsely admire and extol the powers of the human mind, we do not search for its
real helps.”
— Novum Organum: Aphorisms [Book One], 1620, Sir Francis Bacon
5. What is the quality of the current
evidence from observational analyses?
August2010: “Among patients in the UK
General Practice Research Database, the
use of oral bisphosphonates was not
significantly associated with incident
esophageal or gastric cancer”
Sept2010: “In this large nested case-control
study within a UK cohort [General
Practice Research Database], we found a
significantly increased risk of oesophageal
cancer in people with previous
prescriptions for oral bisphosphonates”
6. What is the quality of the current
evidence from observational analyses?
April2012: “Patients taking oral
fluoroquinolones were at a higher risk of
developing a retinal detachment”
Dec2013: “Oral fluoroquinolone use was
not associated with increased risk of
retinal detachment”
7. What is the quality of the current
evidence from observational analyses?
BJCP May 2012: “In this study population,
pioglitazone does not appear to be significantly
associated with an increased risk of bladder
cancer in patients with type 2 diabetes.”
BMJ May 2012: “The use of pioglitazone is
associated with an increased risk of incident
bladder cancer among people with type 2
diabetes.”
8. What is the quality of the current
evidence from observational analyses?
Nov2012: FDA released risk
communication about the bleeding risk of
dabigatran, based on unadjusted cohort
analysis performed within Mini-Sentinel
Dec2013: “This analysis shows that the
RCTs and Mini-Sentinel Program show
completely opposite results”
Aug2013: “However, the absence of any
adjustment for possible confounding and
the paucity of actual data made the
analysis unsuitable for informing the care
of patients”
10. BMJ study design choices
• Data source: General Practice Research Database
• Study design: Nested case-control
• Inclusion criteria: Age > 40
• Case: cancer diagnosis between 1995-2005 with 12-months of
follow-up pre-diagnosis
• 5 controls per case
• Matched on age at index date, sex, practice, observation period
prior to index
• Exposure definition: >=1 prescription during observation period
• “RR” estimated with conditional logistic regression
• Covariates: smoking, alcohol, BMI before outcome index date
• Sensitivity analyses:
• exposure = 2+ prescriptions
• covariates not missing
• time-at-risk = >1 yr post-exposure
11. • In the design of observational studies we rely heavily
on “clinical judgment”
hubris
• Even worse, we do so with very limited feedback
• operating characteristics?
• Like early days of lab testing – “trust me, I measured
it myself”
13. Range of estimates across high-dimensional propensity
score inception cohort (HDPS) parameter settings
Relative risk
True -
False -
False +
True +
Parameter settings explored in OMOP:
Washout period (1): 180d
Surveillance window (3): 30 days from
exposure start; exposure + 30d ; all time
• Each row represents a drug-outcome
from exposure start
pair.
Covariate eligibility window (3): 30
• The horizontal span reflects the
days prior to exposure, 180, all-time
range of point estimates observed
pre-exposure
across the parameter settings.
# of confounders (2): 100, 500
• Ex. Benzodiazepine-Aplastic
covariates used to estimate propensity
anemia: HDPS parameters vary in
score
estimates from RR= 0.76 and 2.70
Propensity strata (2): 5, 20 strata
Analysis strategy (3): Mantel-Haenszel
stratification (MH), propensity score
adjusted (PS), propensity strata
adjusted (PS2)
Comparator cohort (2): drugs with
same indication, not in same class; most
prevalent drug with same indication,
not in same class
13
14. Range of estimates across univariate self-controlled
case series (USCCS) parameter settings
True -
False -
False +
True +
USCCS Parameter settings explored in
OMOP:
Condition type (2): first occurrence or all
occurrences of outcome
Defining exposure time-at-risk:
Days from exposure start (2): should we
include the drug start index date in the
period at risk?
Surveillance window (4):
30 d from exposure start
Duration of exposure (drug era start through
drug era end)
Duration of exposure + 30 d
Duration of exposure + 60 d
Precision of Normal prior (4): 0.5, 0.8, 1, 2
• For Bisphosphonates-GI Ulcer hospitalization,
USCCS using incident events, excluding the first day
of exposure, and using large prior of 2:
• When surveillance window = length of
exposure, no association is observed
• Adding 30d of time-at-risk to the end of
exposure increased to a significant RR=1.14
Relative risk 14
21. Smith et al. 2011 study design and
results
• Data source: Administrative claims from health insurance board of Quebec
• Study design: Cohort
• Exposure: all patients dispensed >=30d of therapy, 180d washout
• Unexposed cohort: 2 patients per exposed, matched by age, gender, and
region, with no tuberculosis therapy
• Time-at-risk: Length of exposure + 60 days
• Events: Incident hospital admission for noninfectious or toxic hepatitis
• “Event ratio” estimated with conditional logistic regression
• Covariates: prior hospitalization, Charlson score, comorbidities
22. Receiver Operating Characteristic
(ROC) curve
False positive rate (1-Specificity)
Sensitivity
• ROC plots sensitivity vs.
false positive rate
• Rank-order all pairs by
RR from largest to
smallest
• Calculate sensitivity and
specificity at all possible
RR thresholds
Isoniazid (RR=4.04):
Sensitivity = 4%
Specificity = 98%
• Area under ROC curve (AUC)
provides probability that method
will score a randomly chosen true
positive drug-outcome pair higher
than a random unrelated drug-outcome
pair
• AUC=1 is perfect predictive model
• AUC=0.50 is random guessing
(diagonal line)
• Cohort method on MDCR:
AUC = 0.64
23. Strategies to improve predictive
accuracy
• Stratify results by outcome
• Tailor analysis to outcome
• Restrict to sufficient sample size
• Optimize analysis to the data source
24. Performance after applying these
strategies
Positives: 19
Negatives: 41
• Restricting to drugs with sufficient sample
• Increased prediction comes as tradeoff with
False positive rate (1-Specificity)
Sensitivity
AUC=0.92 AUC=0.76
AUC=0.84 AUC=0.86
Positives: 51
Negatives: 28
Positives: 30
Negatives: 48
further increased AUC for all outcomes, but
the degree of change varied by outcome
Positives: 22
Negatives: 47
fewer drugs under surveillance
• Self-controlled cohort design continue to be
optimal design, but specific settings
changed in all outcomes
25. Performance across methods, by
database
Data source
AUC for pairs with MDRR<=1.25
• All self-controlled designs (OS, ICTPD, SCCS) are consistently at or near the top of
performance across all outcomes and sources
• Cohort and case-control designs have comparable performance, consistently lower
than all self-controlled designs
• Substantial variability in performance across the optimal settings of each method
26. Good performance?
• …it all depends on your tolerance of false
positives and false negatives…
• …but we’ve created a tool to let you decide
http://elmo.omop.org
27. Revisiting clopidogrel & GI bleed
(Opatrny, 2008)
OMOP, 2012 (CC: 2000314, CCAE, GI Bleed)
Relative risk: 1.86, 95% CI: 1.79 – 1.93
Standard error: 0.02, p-value: <.001
31. Evaluating the null distribution?
• Current p-value calculation assumes that you have an
unbiased estimator (which means confounding either
doesn’t exist or has been fully corrected for)
• Traditionally, we reject the null hypothesis at p<.05 and
we assume this threshold will incorrectly reject the null
hypothesis 5% of time. Does this hold true in
observational studies?
• We can test this using our negative controls
32. Ground truth for OMOP 2011/2012
experiments
Positive
controls
Negative
controls Total
Acute Liver Injury 81 37 118
Acute Myocardial Infarction 36 66 102
Acute Renal Failure 24 64 88
Upper Gastrointestinal Bleeding 24 67 91
Total 165 234 399
Criteria for negative controls:
• Event not listed anywhere in any section of active FDA structured product label
• Drug not listed as ‘causative agent’ in Tisdale et al, 2010: “Drug-Induced
Diseases”
• Literature review identified no evidence of potential positive association
33. Negative controls & the null
distribution CC: 2000314, CCAE, GI Bleed
clopidogrel
34. Negative controls & the null
distribution CC: 2000314, CCAE, GI Bleed
55% of these
negative controls
have p < .05
(Expected: 5%)
40. p-value calibration plot
CC: 2000314, CCAE, GI Bleed
p < .05 55%
Calibrated p < .05 6%
clopidogrel:
RR 1.9 (1.8 – 1.9)
p <.001
Calibrated p .30
clopidogrel
41. p-value calibration plot
CC: 2000314, CCAE, GI Bleed
This analysis failed to
reject the empirical null
… but we know
clopidogrel causes GI
bleeding (it’s a positive
control)
clopidogrel
42. p-value calibration plot
Optimal method: SCCS:1931010, CCAE, GI Bleed
p < .05 33%
Calibrated p < .05 9%
clopidogrel:
RR 1.3 (1.2 – 1.3)
p <.001
Calibrated p .01
43. Applying case-control design to
negative controls in real data
45% of the CIs of
negative controls
contain 1
(Expected: 95%)
44. ‘True RR’ – injected signal size
Coverage probability
Coverage probability by effect size
45. Conclusions
• Using an empirically driven, high throughput
approach to study design, a risk identification
system can perform at AUC>0.80
• Traditional p-values and confidence intervals
require empirical calibration to account for bias in
observational studies
• Advancing the science of observational research
requires an empirical and reproducible approach
to methodology and systematic application
• Predictive models face similar issues
46. Introducing OHDSI
• The Observational Health Data Sciences and
Informatics (OHDSI) program is a multi-stakeholder,
interdisciplinary collaborative to
create open-source solutions that bring out
the value of observational health data through
large-scale analytics
• OHDSI has established an international
network of researchers and observational
health databases with a central coordinating
center housed at Columbia University
47. OHDSI’s vision
OHDSI collaborators access a network of 1
billion patients to generate evidence about all
aspects of healthcare. Patients and clinicians
and other decision-makers around the world use
OHDSI tools and evidence every day.
48. To go forward, we must go back
“What aspects of that association should we especially
consider before deciding that the most likely
interpretation of it is causation?”
• Strength
• Consistency
• Temporality
• Plausibility
• Experiment
• Coherence
• Biological gradient
• Specificity
• Analogy
Austin Bradford Hill, “The Environment and Disease:
Association or Causation?,” Proceedings of the Royal
Society of Medicine, 58 (1965), 295-300.
49. Introducing HOMER
• Health Outcomes and Medical Effectiveness Research
(HOMER) system
• Live, interactive evidence exploration system with fully
functional implementations of all of the components of Sir
Austin Bradford Hill’s viewpoints for risk identification and
assessment, plus some additional components designed by
the OMOP team
50. HOMER implementation of Hill’s viewpoints
Strength Plausibility
Consistency
Temporality
Experiment
Coherence
Analogy
Biological
gradient Specificity
Comparative
effectiveness
Predictive
modeling
51. CCD
• CCD is a tool for performing massive
regularized regressions (Poisson, linear,
logistic, survival)
– > 1+ million variables
– > 10+ million rows
• Useful for
– Data-driven propensity score estimation
– Building predictive models
– Large scale outcome models (e.g. SCCS)
52. Installing the OHDSI packages
• Install RTools:
http://cran.r-project.org/bin/windows/Rtools/
• In R:
install.packages(“devtools”)
library(devtools)
install_github(“ohdsi/DatabaseConnector”)
install_github(“ohdsi/SqlRender”)
install_github(“ohdsi/CCD”)
install_github(“ohdsi/CohortMethod”)
53. Join us on the journey
• Observational data has tremendous potential value for
healthcare, but systematic application of large-scale
analysis is required to unlock that value
• OHDSI has established a collaborative led by people
and expertise necessary to execute on this ambitious
plan to develop and apply solutions to generate
reliable evidence that can transform medical decision
making
• With your support, together we can make this vision a
reality
http://ohdsi.org
54. Large-scale analytics can help reframe the
patient-level prediction problem
Given a patient’s clinical
observations in the
past….
…can we predict
outcomes for that
patient in the future?
Outcome: Stroke
Age
Gender
Race
Location
Drug 1
Drug 2
…
Drug n
Condition 1
Condition 2
…
Condition n
Procedure 1
Procedure 2
…
Procedure n
Lab 1
Lab 2
…
Lab n
0 76 M B 441 0 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1
1 77 F W 521 1 0 0 1 0 0 1 1 0 0 0 0 1 0 0 0
1 96 F B 215 1 1 1 0 1 0 1 1 1 0 0 0 1 1 0 1
1 76 F B 646 0 1 0 0 1 0 1 1 0 0 0 0 1 0 1 0
0 64 M B 379 0 0 1 1 1 1 0 1 1 1 0 1 0 0 0 0
1 74 M W 627 0 1 1 1 0 0 1 1 0 1 1 1 1 0 0 1
1 68 M B 348 0 0 0 1 0 0 1 0 1 1 1 0 1 1 0 1
Demographics All drugs All conditions All procedures All lab values
55. Tools for Large-Scale Regression
BBR/BMR
2005
bayesianregression.org
logistic, multinomial
L1, L2 regularization
sparse millions of predictors
hierarchical, priors, autosearch
stable
BXR bayesianregression.org
cleaner
BOXER
online logistic regression
CCD
CYCLOPS
bsccs.googlecode.com
logistic, conditional logistic,
multinomial, Poisson, Cox,
ParamSurv, least squares
L1, L2 regularization
sparse millions of predictors
imputation
CPU, GPU
56. PLATO
Patient-Level Assessment of Treatment Outcomes
• Predictive models that assess the probability of a patient
experiencing any outcome following initiation of any
intervention, given his or her personal medical history
• Front-end prediction browser
• Challenge: ~25K interventions X ~1K outcomes = 25M
models
57. Sparse Coding Relational Random Forests
(>-30, appendectomy, Y/N):
in the last 30 days, did the patient have an appendectomy?
(<0, max(SBP), 140):
at any time in the past did the patient’s SBP exceed 140 mmHg?
(<-90, rofecoxib, Y/N):
in the time period up to 90 days ago, did the patient have a
prescription for rofecoxib?
(>-7, fever, Y/N):
in the last week, did the patient have a fever?
Bayesian Hierarchical Association Rule Mining
Shahn et al.
McCormick et al.
• Goal: Predict next event in current
sequence given sequence database
• So far, successful application to RCT
data
Bayesian List Machine
Rudin et al.
58. Conclusions
• Using an empirically driven, high throughput
approach to study design, a risk identification
system can perform at AUC>0.80
• Traditional p-values and confidence intervals
require empirical calibration to account for bias in
observational studies
• Advancing the science of observational research
requires an empirical and reproducible approach
to methodology and systematic application
• Predictive models face similar issues