• SxP Organizing Symposium in 2020 WCoP: Model Averaging and Estimands
    • SxP members Presented in PSI Scientific Committee Webinar - Longitudinal Modeling: Time to take the next Step? - November 18, 2019:
      • Presentation slides can be found here.
    • SxP in ACP10:
      • SxP Session Recording: "You say 'TOMATO', I say 'TOMAHTO'"opening_logo.png
        • The Case of the Crisis of Confidence: Jim Rogers sued Eric Jordie for improperly labelling prediction intervals in his VPC as confidence intervals. Judge Kowalski found Dr. Jordie guilty with extenuating circumstances, given that the ISoP guidance on model evaluation is what steered him wrong! Eric was held in contempt of court for attitude problems
        • The Case of the Pharmacometrician’s Pitiful Planning: France Mentre sued Jin Jin for not providing pharmacometric input into the sample size of a clinical trial for which PK was a key component. This case was decided by a jury, the members of which were randomly selected from the audience. While the jury (led by Matt Zierhut) found Jin not guilty, they suggested that she visit a clinical site participating in the Pop PK substudy to help her appreciate the challenges sites face, and to attend a seminar on persuasive presentations with Pete Bonate to help her convince the team to get the necessary samples next time.
        • The Case of the Extrapolated Prediction: Lei Nei sued Chao Liu for making drug development decisions based solely on extrapolated predictions. Specifically, Chao had recommended a clinical dose for Formulation Y based on an exposure-response model for Formulation X and a PK model for Formulations X and Y. Judge Kowalski dismissed the case given then the Chao was using well-established pharmacometric approaches. However, he reminded Chao to explicitly acknowledge his assumptions when making extrapolated predictions.
        • The Case of the Not So Confirmatory Conclusion: Brian Smith sued Nag Chematuri for overstating the significance of covariate effects in a population pharmacokinetic analysis (using p-values for exploratory analysis). Nag had tested 34 covariates on all of the structural parameters in his PK model, without any adjustment for multiple testing. Judge Kowalski found Nag guilty, but the case was appealed to the Supreme Court where Justice Brian Corrigan overruled with a resounding GUILTY verdict…. When Judge Kowalski pointed out that his verdict was actually the same, Justice Corrigan got a big laugh by saying “I know, but I’ve been wanting to overrule you for 25 years!!” 
        • The full scripts and the slide deck can be found here.
      • Judges in SxP Session "You say 'TOMATO', I say 'TOMAHTO'"
    • First SxP Student Award given to Sihang Liu (Buffalo Univ)
    • Announcement of the MBMA subgroup, by Matt Zierhurt

    • Presentations from SxP session in 2019 ISBS have been uploaded to the library. You can find them here.
      Presentations from SxP Session in 2019 JSM have been uploaded to the library. You can find them here.

      The INSERM research group led by France Mentre, in Paris, held the first ISoP SxP SIG Local event, on the 11th of July.

    The program was the following:

    • 14:00-14:40 Prof Andrew Gelman (Columbia Univ.) - “We've Got More Than One Model: Evaluating, comparing, and extending Bayesian predictions” 
    • 14:40-15:20 Dr. Sebastien Benzekry (INRIA) - “Mechanistic modeling of metastasis: cancer at the organism scale”
    • 15:20-15:40 Witold Wiecek (Certara) - “Bayesian meta-analysis model for chemical risk assessment”
    • 15:40-16:00 Francois Riglet (INSERM Paris Univ.) – “Bayesian individual dynamic predictions of biomarkers and risk of event in joint modelling (with uncertainty): a comparison between Stan, Monolix and NONMEM”
    • Coffee/tea/macarons break
    • 16:30-17:10 Prof Rob Bies (Buffalo Univ.) - “A hybrid genetic algorithm for NONMEM structural model optimization”
    • 17:10-17:50 Mike Smith (Pfizer) - “Dose-response, small data & big decisions”
    • 17:50-18:10 Shan Pan (St John's Institute of Dermatology) - “Real-world clinical effectiveness and utility of ustekinumab in adults with moderate to severe psoriasis: statistical and pharmacometric perspectives”

    Prof France Mentre and Mike K Smith (Pfizer), co-chairs of SxP SIG, welcomed the audience to this first local event of this ASA/ISoP SIG. They handed the floor to Prof Andrew Gelman (Columbia Univ) who gave an opinion piece (no slides) on how indeed all models are wrong but some are useful and certainly none are ridiculous. Then, Dr Sebastien Benzekry (INRIA) illustrated how models of partial differential equations can help understand and predict tumour metastasis development in breast and lung cancer. These two invited talks were followed by submitted abstracts: Dr Wiecek Witold (Certara and LSE) on uncertainty factor meta-analysis and PhD student Francois Riglet (INSERM) on individual dynamic predictions. After a short macaroon break, Prof Robert Bies (Buffalo Univ) presented the Single Objective Hybrid Genetic Algorithm and its implementation in a Shiny App using nonmem. Then, Mike K Smith, guided the audience through 4 steps in drug development where statisticians and pharmacometricians work better together.  Finally, PhD student Shan Pan (St John's Institute of Dermatology) illustrated on ustekinumab how statisticians and pharmacometricians can benefit from working together.  You can find presentations here.

    thumbnail image
    Our New SxP Co-Chair, Mike Smith, has won the PSI/RSS Statistical Excellence Awards in the Pharmaceutical Industry!
    As reflected in the nomination letter, this award is related to Mike's work on SxP:

    "...Your entry 'Excellence in research around population modelling in drug development' was recognized for pioneering practical Bayesian methods with drug development in the context of optimising decision making within trials and modelling dose response with pharmacokinetic/pharmacodynamic modelling..."

    The award is presented today, June 4th, 2019, at the PSI conference in London.

    Congratulations, Mike!

  • 2019 ACOP10 Pictures :
 Dr. France Montre in Meet SxP SIG Lunch:

  • 2018 Pharmacometric Programming Webinar: 

    The American Statistical Association (ASA) Section for Statistical Programmers and Analysts (SSPA) and the Statistics and Pharmacometrics Interest Group (SxP) jointly sponsored a webinar on Pharmacometric Programming. The webinar will take place on Tuesday, Jan 30th from 12:00-1:30 pm. The description and registration instructions are listed below. Please note that multiple persons are encouraged to view each registered connection (for example, by projecting the webinar in a conference room).

    Title: Pharmacometric Programming
    Presenters: Amit Roy, Jing Su, Neelima Thanneer, and Jeffry Florian
    Date and Time: Tuesday, January 30, 2018, 12:00 p.m. – 1:30 p.m. Eastern time
    Registration Deadline:
    Friday, January 26, at 12:00 p.m. Eastern time

  • ASA Joint Statistical Meeting (JSM) 2017: July 29 – August 3 2017 in Baltimore, WA. Topic contributed papers by our interest group on Pharmacometric Programming: Thu. August 3rd, 2017, 8:30 AM - 10:20 AM

  • Clinical Pharmacology Workshop Series 2017: August 9 – 11 2017 in Salt Lake City, Utah, USA. By University of Utah, Department of Pediatrics, Division of Clinical Pharmacology. 

  • CENISBS: Joint Conference on Biometrics & Biopharmaceutical Statistics: 28 August – 1 September 2017 in Vienna, Austria 

  • PODE: 12th Workshop on Population Optimum Design of Experiments (PODE) 2017, Friday, 8 September 2017, the Biostatistics and Pharmacometrics department of Novartis Pharma.
    One of the highlights will be our Steering Committee Member, Jose Pinheiro's talk, on “Exposure-response modeling for dose selection under model uncertainty: Extending the MCP-Mod approach”. This goes beyond the typical PK sampling time point optimization and exemplifies the relationship between study design aspects and optimal design considerations.

  • SxP Proposed session accepted by ACoP8 (October 15-18, 2017 in Fort Lauderdale, FA, USA): Integrating quantitative disciplines - Making model-informed discovery and drug development (MID3) work in practice, proposed by Mark K. Smith (Pfizer).

    To truly deliver model-informed discovery and drug development (MID3), we need to integrate information across a wide range of quantitative disciplines - systems pharmacology, clinical pharmacology, pharmacometrics, statistics, epidemiology, decision scientists. Yet there is a danger that each discipline works in isolation, developing models and making inferences. Models and information sharing struggles to cross between quantitative groups. Assumptions are not well captured, articulated or shared in ways that make model scope and limitations clear. Instead of accumulating knowledge and understanding across the quantitative groups, we get mistrust, rework and a myopic view. But by embracing the principles of MID3, employing good practice, and finding suitable ways to share and build on knowledge which accumulates across disciplines we can work more effectively to bring a complete picture of drug action to development of new medicines and drug regimens, and to use that information to plan and design efficient studies. In this session we will look at how the quantitative disciplines might work together effectively, break down silos, work together on models, effectively share information, inform design, make better inferences and decisions.