Add ability to return the inferred data or longitudinal flagging

[dewittpe]

The most common question I get about the longitudinal flags is generally something along the lines of "how long should comorbidities be carried forward?"

I don't have a good answer to that question because some comorbidities are life long and others might not be. More research is needed

To support the research, medicalcoder needs to provide the inferred longitudinal dataset used for the flagging of comorbidities.

Initial thought is to modularize the comorbidities() function and allow users to call some of the functions, or have the return object dependent on arguements passed to comorbidities()

[dewittpe]

An other thought is to have the return object from comorbidities() be a list with elements for current, cumulative, subconditions, metadata. with this structure it would also be possible to add an element with a data.frame indicating when a cumulative flag is present in conjunction with a current flag, between two current flags, only following a current flag, when a cumulative flag is reported on subsequent encounters after ICD code(s) for a condition are reported but the condition is not flagged. An example for the last case is K76.0 for mild liver disease under AHRQ Elixhauser 2022--today methods. A mild liver disease is only flagged if present on admission. If K76.0 is non-poa on encounter 1, then mlid liver disease would not be flagged on encounter 1 under flag.method = 'current' nor flag.method = 'cumulative'. However, encounters 2, 3, ... would have a flag under flag.method = 'cumulative'.