jaygee-on-github / the-aphrc-lshtm-mubas-idsr-project Goto Github PK

• ATLAS supports the cohort method design which may be comparable to the statistical analysis you specify for Objective 1. Here I will need additional guidance from you to determine if, in fact, the two approaches are comparable.

• IDSR PUIs fall into two categories — community and institutional. I am not sure how you want to work with two study populations in the design for Objective 1

• In Objective 1 you specify two “exposures” or, again, interventions — Vaccinated and Unvaccinated. Vaccinated includes one or more doses, regardless of the vaccine. You may want to consider an analysis with three exposures — Vaccinated, Partially Vaccinated and Unvaccinated where we are sensitive to dosage requirements by vaccine type.

• In terms of secondary outcomes, both hospitalization and the risk of a severe COVID-19 disease experience are off the table. The IDSR does not collect hospitalization information as a rule although individual countries may have augmented the IDSR to capture this information. Likewise the IDSR doesn't collect condition information among PUIs who test positive as a rule.

• In term of secondary outcomes, the IDSR, as a rule, does collect death information. So death can be its own outcome with its own logistic regression and odds ratio.

• Control for confounders maybe uses a strategy that is a little different (or not) in the cohort method design and the statistical analysis you specify for Objective 1. We can discuss and you will need to judge.

• Objective 2 is currently out of scope because so far we have not augmented the IDSR with a questionnaire. In fact, depending on the country, a subpopulation of the institutional PUIs are health care workers. They could be retrospectively questioned for barriers and enhancers.

• Objective 3 is out of scope.

The source specimen table was mapped to the OMOP CDM SPECIMEN and OBSERVATION tables. The source lab results table was mostly mapped to to the OMOP CDM MEASUREMENT table.

With MEASUREMENTs the best practice in the OMOP CDM is to use concepts from LOINC for the measurement_concept_id. The measurement_concept_id for first the PCR test and then the antigen test are as follows:

The selection of these measurement_concept_ids was based on the following considerations:

• As a rule the at home rapid antigen test was not available early on in the pandemic. Instead there were sites (clinic, hospitals, commercial labs) that conducted PCR tests.
• On occasion these sites might conduct an antibody test.
• As a rule the PCR test was considered "the gold standard".
• As a rule nasal swabs were the source of both the PCR tests at first and then the rapid antigen test later. The occasional antibody test used blood.

Here is the measurement_concept_id for the occasional antibody test:

The admission is mapped to one or more CONDITION_OCCURRENCEs, depending on whether presenting symptoms were captured or not.

The CONDITION_OCCURRENCE corresponding to the suspected disease (COVID-19) has a condition_concept_id corresponding to "Suspected COVID-19" in ATHENA (37311060). This is a concept from the SNOMED vocabulary. This CONDITION_OCCURRENCE has a condition_status_concept_id corresponding to "Admission diagnosis" in ATHENA (32890). This concept comes from the Condition Status vocabulary. This same CONDITION_OCCURRENCE takes a condition_type_concept_id in line with its "provenance" (32809). This is the concept for a "Case Report Form" in the Type Concept vocabulary.

Alternatively, the condition_concept_id can be "At an increased risk to exposure by Severe acute respiratory system corona virus 2".

• To adjust for differences between the two treatment groups in the OHDSI data analysis work bench, several adjustment strategies can be used, such as stratification, matching, or weighting by the propensity score, OR by adding baseline characteristics to the outcome model. It appears in your analysis, you are adding baseline characteristics to a logistic regression model and you are not using the propensity score in any way, shape or form.

  • Recall that in a propensity score-adjusted observational study, we estimate the probability of a patient receiving the target treatment based on what we can observe in the data on and before the time of treatment initiation (irrespective of the treatment they actually received).

  • The PS can be used in several ways including matching target subjects to comparator subjects with similar PS, stratifying the study population based on the PS, or weighting subjects using Inverse Probability of Treatment Weighting (IPTW) derived from the PS.

  • In one-on-one PS matching, we find one or more matched patients that received the comparator but had the same prior probability of receiving the target. Then we compare the outcome for the target patient with the outcomes for the comparator patients within each of these matched groups

  • Now consider the alternative which adds baseline characteristics to a logistic regression model in ATLAS

  • In this approach ATLAS does NOT identify confounders by stratifying a vaccine effectiveness statistic on candidate confounders. Instead ATLAS tries the regression with all the possible characteristics. Indeed, there is no "selection". ATLAS is simply driven by the available data.

  • Note that "Propensity score analysis is overall a more favorable approach than traditional regression analysis when estimating causal effects using observational data." See here.

• This discussion suggests we create two emulated clinical trial templates — a “traditional" one that uses the outcome model only to identify confounders and a template that specifies how to use propensity scores either alone or in conjunction with the outcome model. We might call this template the “advanced” emulated clinical trial template based on research that compares the two approaches.

The IDSR Immediate Case-based Reporting Form used in the African Region does not actually capture they type of vaccine a person received:

This is consequential when it comes to COVID-19 because, in principle, it is necessary to know not just the number of doses but also the vaccine type in order to determine if a vaccinated COVID-19 IDSR case is either fully or partially vaccinated based on the number of vaccine does received and the date of last vaccination.

Under the circumstances, we considered a couple of strategies.

In the first strategy we just assume someone is fully vaccinated if they have received two or more doses even though certain vaccines like Janssen COVID-19 Vaccine (Johnson & Johnson) are "one-and-done".

In the second strategy it is possible to augment the IDSR with country-specific information provided by COVAX:

From the WHO Africa COVID-19 dashboard

Recall that COVAX had a hand in procuring and distributing COVID-19 vaccine doses across the African region. This dashboard includes both the COVAX procurements and the non-COVAX procurements.

In a proof-of-concept INSPIRE adopted the second strategy: we assigned a vaccine type probabilistically to each PERSON who received one or more vaccine doses. Note that in certain countries like Malawi the WHO CRF was amended to include the type of vaccine. This led us to think that in some countries we would and in other countries we wouldn't have to use the COVAX data during analysis.

With vaccine doses reported, using the IDSR and COVAX, INSPIRE created a synthetic source table that looked like this:

During the ETL each row in the source data was used to create a CDM DRUG_EXPOSURE record. Just like with the other tables in the OMOP CDM, a [domain]_concept_id needs to be provided. With a DRUG_EXPOSURE, this would be a drug_concept_id. It comes from one of the standard vocabularies hosted and managed by ATHENA. Here the drug_concept_id identifies the specific COVID-19 vaccine a person received. These concepts come from the standard vocabulary which is actually a set of specific vocabularies that include SNOMED-CT, LOINC, ICD-10, RxNorm and others. One of the "others" is CVX which hosts vaccine codes for vaccines approved for use in the United States as well as vaccines used internationally.

Also note that a DRUG_EXPOSURE also has a drug_type_concept_id which provides the provenance for this information is the Case Report Form.

How were these coded during the ETL? @tathagatabhattacharjee, what do you think? I don't have Michael's address but we will try to add it.

Recall that the outcome corresponds to whether the initial classification was "confirmed" or not. In the OMOP CDM this amounts to another CONDITION_OCCURRENCE. This one follows in time the admission CONDITION_OCCURRENCE(s). It has a condition_concept_id corresponding to "COVID_19" in ATHENA (37311061). This concept comes from the SNOMED vocabulary. It subsumes many more specific COVID-19 related conditions, but not "Suspected COVID-19":

The outcome CONDITION_OCCURRENCE also takes a condition_status_concept_id and a condition_type_concept_id. The condition_status_concept_id corresponds to "Confirmed Diagnosis" (32893). "Confirmed Diagnosis" is a concept in the Condition Status vocabulary. This CONDITION_OCCURRENCE takes a condition_type_concept_id in line with its "provenance" (32809). This is the concept for a "Case Report Form" in the Type Concept vocabulary.