Stats: How do you analyze safety data
(January 22, 2008). Someone on the MedStats email discussion group asked
about how to analyze adverse event data. He noted that adverse event data is
not one of the primary or secondary outcome measures, and wondered if it
would be appropriate to provide statistical analysis of this data. Adverse
events (and safety data in general) represent a special type of analysis that
does not fit in well with the listing of primary/secondary outcomes. The main
reason for this is the number of possible adverse event categories is very
broad and it is not always possible to anticipate in advance what type of
adverse events are of greatest interest.
Stats: A new and simple approach
for monitoring safety data (November 18, 2007). Many hospitals
administrators collect safety data, and for the most part this data is not
analyzed well. The people who collect the data are well-meaning, but the
simplistic tables and graphs that they use are typically unable to reveal
important trends and patterns in the data. Much of the safety data represents
a description of events (usually bad events) that occur. The question that
always seemed to be on their minds was: is there a sudden surge of events
that we need to take action on?
Stats: The pros and cons of control
charts versus data mining (November 17, 2007). In a talk I gave in
December 2006, I highlighted how in the analysis of adverse event data,
control charts can augment more complex statistical tools like data mining.
Here's a summary of the pros and cons of using control charts.
Stats: Monitoring adverse events
during peritoneal dialysis (November 15, 2007). One of the doctors I was
working with had an interesting data set examining adverse events in patients
with peritoneal dialysis. These patients start treatment with peritoneal
dialysis on a specific day and are followed until they stop this treatment.
There were two adverse events examined: exit site infections, and
peritonitis. Although I ran several complex analyses on this data set, I
thought it might be useful to look at a simpler approach to monitoring the
frequency of adverse events using control charts.
Stats: NNH talk update (November 12,
2007). Last year, I gave a talk for PharmaIQ about continuous monitoring
of the number needed to harm. I want to update this talk for a second
audience in December.
Stats: Tracking
adverse events during kidney biopsy, Part 2 (April 5, 2007). This is a
major revision of an earlier weblog entry. I have been helping a colleague
who is interested in monitoring the safety of kidney biopsy events. He was
kind enough to let me use his data set on my web pages in order to illustrate
some new methods for monitoring adverse events. This data set allows you to
see some examples of the use of control charts to track adverse events. Here
is the raw data.
Stats: Tracking adverse
events during kidney biopsy (March 14, 2007). I have been helping a
colleague who is interested in monitoring the safety of kidney biopsy events.
He was kind enough to let me use his data set on my web pages in order to
illustrate some new methods for monitoring adverse events. This data set
allows you to see some examples of the use of control charts to track adverse
events.
Stats: Two talks for PharmaIQ
(September 19, 2006). I may be giving a couple of talks for for PharmaIQ,
a division of the International Quality & Productivity Center (IQPC). The
first has the title "Signal Detection Strategies for Paediatric Treatments"
and the second has the title "Control charts for continuous monitoring of the
number needed to harm."
Stats: Continuous
monitoring of the number needed to harm (September 2, 2006). The
continuing review of clinical trials has to address "good news" issues. Does
one arm of the study show substantially better efficacy? Does one arm of the
study have a significantly better safety profile? There are rigorous and well
accepted approaches for determining partway through a clinical trial whether
one arm has a greater proportion of cured patients or a smaller proportion of
harmed patients. Continuing review also has to address "bad news" issues. Is
the study falling behind schedule on its planned enrollment rates? Are
patients dropping out of the study at an alarming rate? Are certain adverse
drug reactions occurring at an unexpected rate? The analysis of "bad
news" issues is more poorly developed. Often decisions about these issues are
based on subjective opinions and ad hoc rules. Statistical process control
charts and Bayesian statistical methods offer an approach to treat on-going
review of rates not tied directly to an efficacy or safety comparison.
Stats: Possible sources of
funding for my grant (July 6, 2006). The NIH has a Request for
Application (RFA) titled Research on Research Integrity (R01). The full text
of this announcement is on the web at grants1.nih.gov/grants/guide/rfa-files/RFA-NR-07-001.html.
The goal of this RFA is to foster empirical research on research
integrity. The sponsoring programs are particularly interested in research
that will provide clear evidence (rates of occurrence and impacts) of
potential problems areas as well as societal, organizational, group, and
individual factors that affect, both positively and negatively, integrity in
research. Applications must have clear relevance to biomedical, behavioral
health sciences, and health services research.
Stats: Applications of the CUSUM
chart (June 20, 2006). I am interested in investigating the use of CUSUM
charts in monitoring accrual rates, drop out rates, and adverse event rates
in a clinical trial. Some references which I might cite in a literature
review are listed here.
Stats: Seminar on control charts
and adverse events (June 5, 2006). I took some time to expand my May 30,
2005 weblog entry on accrual rates and developed a seminar which I will
present to the Statistics journal club at KUMC today. The handout for this
talk combines that weblog entry with a brief tutorial on quality control. I
received some valuable feedback.
Stats: Upcoming talks about
control charts (May 25, 2006). I am working on some ideas for a grant to
use control charts to track adverse events in clinical trials. I also
envision the possibility of using control charts as a warning of a sudden
influx of events that may be an early indicator of a bioterrorism event. I
have not fleshed out these ideas very completely yet, but hope to do so soon
in the weblog. While reviewing the upcoming talks at the Joint Statistics
Meeting in Seattle, August 2006, I noticed several interesting talks that
appear to be related to some of the things I might be working on.
Stats: Data mining and drug safety
(May 4, 2006). I am very interested in safety issues, especially in the
continuing review/interim analysis of clinical trials. It turns out that
S-plus is targeting drug safety as a particularly important application of
its data mining modules. Two recent web seminars addressed this topic.
Stats: I want to write a grant
(April 25, 2006). I have been mulling over the idea of writing a research
grant where I am the primary investigator. I have helped lots of other people
write grants, but have never before taken the step of writing a grant myself.
I have a rough idea of the form that this grant would take, but I want to use
this weblog to flesh out these ideas and articulate them more clearly.
Stats: Reporting serious adverse
events (updated February 3, 2006). The FDA held a meeting on March 21,
2005 soliciting opinions about how adverse events should be reported to
Institutional Review Boards (IRBs). Some of the testimony provided to FDA can
be found on the FDA website and in various spots on the Internet, mostly in
PDF format. This is something I have been interested in, but have not had the
time to work up the details. It seems to me that any system for reporting
adverse events has to have information about the accrual of patients into the
study. Here's a simple graph that shows the entry and exit times in a
research study. It's not exactly a study of adverse events reports per se,
but the example is close enough that I can use to illustrate the general
concepts.
Stats: Reporting of adverse
events (August 5, 2005). Most Institutional Review Boards (IRBs) have
difficulty coping with the volume of adverse events that study sponsors
report to them. The FDA held a public meeting about this issue recently, and
some written responses are available as PDF files at the following location:
www.fda.gov/ohrms/dockets/dockets/05n0038/mostrecent.htm.
Stats: Control charts for
monitoring mortality rates (February 11, 2005). One of the trickiest
problems in Medicine is trying to identify whether an unusual trend in
mortality rates is an indication of an incompetent physician, or worse, a
physician who is actively killing patients.
