Multi-Site Analysis for Independent Data Sources

The multi-site analyses included in this suite are intended to be executed against data that are all stored in the same place. However, there may be some instances where the data associated with each site is stored in independent locations. This vignette outlines how the multi-site analysis can be executed in these instances.

After following the instructions to reproduce the analysis, you will also need to change the output_function column to tell the csd_output function which check you executed. Reference the table below for the labels that are associated with each check:

Check Type	output_function
Multi Site, Exploratory, Cross-Sectional	csd_ms_exp_cs
Multi Site, Exploratory, Longitudinal	csd_ms_exp_la
Multi Site, Anomaly Detection, Cross-Sectional	csd_ms_anom_cs
Multi Site, Anomaly Detection, Longitudinal	csd_ms_anom_la

Multi-Site Exploratory Analysis

The process for the exploratory analysis is the same for both the cross-sectional and longitudinal configurations.

First, execute either of the Single Site, Exploratory analyses, configured appropriately for your study, against each data source.

library(conceptsetdistribution)

my_table <- csd_process(cohort = my_cohort,
                        multi_or_single_site = 'single',
                        anomaly_or_exploratory = 'exploratory',
                        time = T / F,
                        ...)

Then, combine these results into a single table with the different sites delineated in the site column.

my_final_results <- my_table1 %>% dplyr::union(my_table2) ... %>%
  dplyr::union(my_table_n) %>%
  dplyr::mutate(output_function = '{see table above}')

Multi-Site Anomaly Detection Analysis

For anomaly detection analysis, start by executing the same steps as the exploratory analysis. Then, you will execute the relevant anomaly detection algorithm against the resulting table. See below for the different processes for cross-sectional and longitudinal analysis.

Cross-Sectional

For a cross-sectional analysis, the compute_dist_anomalies and detect_outliers functions, both available through the squba.gen package, should be executed against your results. Copy the code below, inputting the table you generated.

If using the OMOP CDM, concept_id should be input as the concept column. For the PCORnet CDM, concept_code should be input as the concept column. The p_value can be selected by the user.

# First execute the compute_dist_anomalies function
df_start <- compute_dist_anomalies(df_tbl = my_table,
                                   grp_vars = c('variable', concept_column),
                                   var_col = 'prop_concept',
                                   denom_cols = c('variable', 'ct_denom'))

# Then, use that output as input for the detect_outliers function
df_final <- detect_outliers(df_tbl = df_start,
                            tail_input = 'both',
                            p_input = p_value,
                            column_analysis = 'prop_concept',
                            column_variable = concept_column) %>%
  dplyr::mutate(output_function = '{see table above}')

Longitudinal

For a longitudinal analysis, the ms_anom_euclidean function, available through the squba.gen package, should be executed against your results. Copy the code below, inputting the data you generated.

If using the OMOP CDM, concept_id should be input as the concept column. For the PCORnet CDM, concept_code should be input as the concept column.

df <- ms_anom_euclidean(fot_input_tbl = csd_tbl,
                        grp_vars = c('site', 'variable',
                                     concept_column),
                        var_col = 'prop_concept') %>%
  dplyr::mutate(output_function = '{see table above}')