Question about treatment outcome
Started by Mamel Quelapio, MD on 28 Aug 2008
I am writing to ask a question regarding outcome. I'd like to cite this almost real scenario which is shared by a few patients in our cohort.
A patient missed treatment for 2 consecutive months - 2 months and 5 days to be exact after having been on treatment for 14 months and culture negative since 2nd month of treatment. He was given an outcome of DEFAULTED.
After 2 months, he came back for treatment with a negative smear and culture. He was reentered and given a new patient number. He remained consistently negative.
The question is:
1. Should he be considered a totally new patient and will have to undergo a complete 18 months treatment?
2. Should we let him finish now as cured (his 5th month; his 19th month of being culture-negative considering the first treatment)? IF this is the case, his original outcome stays as DEFAULTED since it will be messy to be changing outcomes which have already been entered in the database and counted as such in a previous analysis. On the other hand, if we say this second enrollment is a CURED we would now have a patient in the cohort who would have only 5 months duration of treatment. BUt we can just make a remark on why this is so.
3. Considering his early conversion and consistent negative status, I don't think he should complete another course of 18 months. But how will programs record this? Do we have guidelines for this?
Thanks in advance for your comments.
Keywords: Clinical Guidelines
Einar Heldal
Dear Mamel,
Thanks again for a good question.
This is my understanding of the guidelines. Whether the patient should start treatment again or not after having defaulted, is of course a clinical question. The doctors have three options:
(1) decide that there is no need to give more treatment, (2) decide to continue the interrupted treatment for a few more months, or (3) decide to start a new full MDR-TB treatment. In the recording and reporting system these options should be managed as follows:
Option (1): treatment outcome defaulted
Option (2): treatment outcome defaulted. The patient does not need to be reentered with a new registration date and category of patient, since he/she is continuing the same treatment,
Option (3): treatment outcome defaulted in the first treatment. The patient should be reentered with a new date of registration and category of patient (previously treated with 2.line drugs, treatment after default).
You may think that it is unfair to have this patient classified as defaulted since he/she may clinically be assessed as "cured". The point of the R&R system is to give a clear picture of how many patients in the cohort had interruptions of treatment more than two months, to follow trends and see if the level is higher in some areas than others.
You may make ADDITIONAL analyses, assessing what actually happened to the patients classified as defaulted, showing that in the end so and so many were cured, etc. as they have done in for instance Latvia. But the basic statistics (as described in the guidelines) should be strictly the same, in order to make fair comparison.
The following is my understanding, but others may think otherwise?
With best regards, Einar
3:36 PM, 28 Aug 2008 | Permalink
Salmaan Keshavjee, MD, PhD
Dear Mamel,
I tend to agree with Einar on this. The clinical question and the reporting question are different.
Clinically, it sounds like he should not do another whole treatment. Personally, I would do 6 to 10 more months and ensure that he remains negative the whole time. This way, he'll have a total of 20 to 24 months of total treatment (with the gap of 2 months).
As for his outcome, unfortunately for paper-work purposes he is a default since he defaulted. We have these cases in Tomsk a lot, and it's frustrating because we as clinicians know that they are cures, but for recording purposes they defaulted from treatment.
One loophole I can imagine: if he converted at month 2 and you have 12 months of negative cultures with 5 consecutive negative sputums at least 30 days apart, you might be able to call him a cure. What do you think Einar?
3:39 PM, 28 Aug 2008 | Permalink
Einar Heldal
Dear Saalman,
There is probably no loophole: The first outcome should be used, and the moment the patient has been without treatment for 2 months (and not already been declared completed treatment by the doctor), he/she should be classified as default, independently of what happens afterwards. Sorry....
PS: It is always normal to have some defaults. One main purpose of the R&R should be to detect areas with less successful outcomes (over time or between areas), so that we can direct our actions and improve performance. So there is no point in "stretching" definitions.
3:42 PM, 28 Aug 2008 | Permalink
Salmaan Keshavjee, MD, PhD
I know it's frustrating Mamel, but unfortunately, these things happen. We can be happy as clinicians knowing in our hearts that the pt is cured, but unfortunately it doesn't make sense to deviate from the standardized reporting. We have this argument frequently in Russia, but have stuck by the WHO reporting.
3:45 PM, 28 Aug 2008 | Permalink
Mamel Quelapio, MD
Dear Einar and Salmaan,
Thanks for your responses. I agree that it can be quite challenging to strike the balance between the clinical and the programmatic ends.
On Einar's options:
In the recording and reporting system these options should be managed as follows:
> Option (1): treatment outcome defaulted: Agree to put this for the first treatment but the question is really on what to do afterwards when he returns.
> Option (2): treatment outcome defaulted. The patient does not need to be reentered with a new registration date and category of patient, since he/she is continuing the same treatment: What happens to the outcome of defaulted? We can put "defaulted then cured". The question is more of how long the treatment should be given on his return.
> Option (3): treatment outcome defaulted in the first treatment. The patient should be reentered with a new date of registration and category of patient (previously treated with 2.line drugs, treatment after default): Agree but the question is really how long the treatment should be given.
This scenario can actually become more complicated if we are faced with:
1. a patient who defaulted and returned after a YEAR again with negative smear and culture but symptomatic. He remains negative for 5 months with clinical improvement. Would be think the same way - stop treatment then or complete at least 18 months considering that it has been a year?
2. a patient who defaulted and came back with a positive smear. This warrants a new treatment regimen. However, it matters what the previous bacteriology was before he defaulted.
It seems to me that there 4 determinants on how to deal with return after Cat IV defaulters. These are:
1. Duration of treatment prior to default
2. Bacteriology prior to default
3. Duration of default
4. Smear status on return
Complex! But in our DOTS Manual of procedures, this sort of complexity has been dealt with using a table that has these different scenarios. I will try to formulate a similar table for MDR-TB patients and request your comments. I believe that with expanding programs, we need to have guidelines no matter how infrequently we think these scenarios would occur. Will get back to you on my proposal.
3:50 PM, 28 Aug 2008 | Permalink
Carole Mitnick, ScD
Thanks, all. I agree with the discussion overall and the importance of separating the guidelines on clinical decision-making from those on reporting.
This was one of the questions that proved most challenging when we first were working on the consensus definitions. We tried to come up with different scenarios depending on when the interruption occurred (first or second year) and on culture/smear status at reinitiation.
Ultimately, I think the decision was that there were no data to support all these varying approaches FOR REPORTING PURPOSES so it was better to stick with one simple approach and collect data that would help inform the revision of definitions. A table that provides additional guidance/insight would be welcome.
3:58 PM, 28 Aug 2008 | Permalink
Irina Gelmanova, MD
Hi Mamel,
Salmaan has already mentioned our challenges with "courses after default" in
Russia
There is a table with recommendations for clinicians how to deal with
returning defaulters from MDR TB treatment. Please, see "PIH Guide to the
Medical Management of Multidrug-Resistant
Tuberculosis" page # 29
http://www.pih.org/inforesources/MDRTB/PIH_Guide_book_final.pdf
However, the table doesn't state exactly how long should be the second
treatment in each case and "24 months of therapy total past intial
conversion" sounds too long for Russian doctors.
Since there is no scientific data about the duration of treatment after
default, "after default" treatment was as long as Russian doctors thought it
was necessary. In some cases the duration of the second treatment was just 6
month because the patient defaulted after 13-15 months of treatment and was
culture negative from the second month of treatment.
For classification purposes we used the following definition of cure for the
second "after default" treatment: "if a patient completed at least 18
months of treatment [duration of treatment before default + duration of
treatment after default] witihn 24-month period and the patient stayed
culture negative for at least 12 last months."
I feel that our definition provides incentives for Tb doctors to search
promptly and treat defaulters.
In your case the patient would be classified as a default at the first
place. If he would restart and stay on treatment for another four-six
months, his short second treatment would be classifed as a "cure"
In case of patients, who defaulted being culture negative but returned a
year after with no evidence of active tb process, Russian doctors would wait
and require smear/culture/x-ray every six month for another two-three years.
Irina
----- Original Message -----
From: "GHDonline (Carol Mitnick, ScD)" <>
To: "Irina Gelmanova, MD" <>
Sent: Thursday, August 28, 2008 3:00 PM
Subject: Re: [Drug Resistant TB] Question about treatment outcome
Reply to: Question about treatment outcome
Thanks, all. I agree with the discussion overall and the importance of
separating the guidelines on clinical decision-making from those on
reporting.
This was one of the questions that proved most challenging when we first
were working on the consensus definitions. We tried to come up with
different scenarios depending on when the interruption occurred (first or
second year) and on culture/smear status at reinitiation.
Ultimately, I think the decision was that there were no data to support all
these varying approaches FOR REPORTING PURPOSES so it was better to stick
with one simple approach and collect data that would help inform the
revision of definitions. A table that provides additional guidance/insight
would be welcome.
--
**Your reply will be sent to the entire community and posted as is.**
See also:
http://www.ghdonline.org/drtb/discussion/question-about-treatment-outcome/
12:07 PM, 2 Sep 2008 | Permalink
Sophie Beauvais
Dear All,
Thank you for your participation.
To continue the discussion, and to follow on Dr. Quelapio’s great suggestion to formulate a table with different scenarios for MDR-TB patients, I saved the table with suggested algorithm for the reinitiation of MDR TB therapy following treatment interruption as an excel spreadsheet in the community: http://www.ghdonline.org/drtb/resource/reinitiation-of-mdr-tb-therapy-following-treatment/. Is there any other table that could be used as a good basis to start a table with scenarios for MDR TB patients?
The other mentioned resource is the WHO Guidelines for the programmatic management of drug-resistant tuberculosis (http://www.ghdonline.org/drtb/resource/emergency-update-2008-guidelines-for-the-programma/).
This document gives the following definition for “defaulted” in treatment outcome for Category IV treatment (page 21): “Defaulted: A Category IV patient whose treatment was interrupted for two or more consecutive months for any reason.”
I believe the form referred to by Dr. Quelapio is the Category IV Treatment Card (Form 1) which is available page 168 to 171 in pdf format, but please reply with correction if this is not accurate.
Thank you.
4:01 PM, 4 Sep 2008 | Permalink