medicaid_utils.adapted_algorithms.py_cdpsmrx package

Submodules

medicaid_utils.adapted_algorithms.py_cdpsmrx.cdps_rx_risk_adjustment module

class medicaid_utils.adapted_algorithms.py_cdpsmrx.cdps_rx_risk_adjustment.CdpsRxRiskAdjustment[source]

Bases: object

classmethod add_age_and_gender_cols(df: DataFrame) → Tuple[DataFrame, List[str]][source]

Adds age & gender related covariates used in CDPS Rx Risk Adjustment model. Returns a dataframe with the below columns:

  • a_under1 - 0 or 1, age<=1

  • a_1_4 - 0 or 1, 1<age<5

  • a_5_14m - 0 or 1, 5<=age<15 male

  • a_5_14f - 0 or 1, 5<=age<15 female

  • a_15_24m - 0 or 1, 15<=age<25 male

  • a_15_24f - 0 or 1, 15<=age<25 female

  • a_25_44m - 0 or 1, 25<=age<45 male

  • a_25_44f - 0 or 1, 25<=age<45 female

  • a_45_64m - 0 or 1, 45<=age<65 male

  • a_45_64f - 0 or 1, 45<=age<65 female

  • a_65 - 0 or 1, 65<=age

Parameters:

dask.DataFrame (df -)

Returns:

  • dask.DataFrame

  • list of str – List of age and gender column names added.

Examples

>>> # Requires a dask DataFrame with 'Female' and 'age' columns
>>> df, lst_cols = CdpsRxRiskAdjustment.add_age_and_gender_cols(df)
>>> 'a_under1' in lst_cols
True
classmethod add_cdps_category_cols(df: DataFrame, lst_diag_cd_col_name: str) → Tuple[DataFrame, List[str]][source]

Adds CDPS category columns

Parameters:

  • df (dask.DataFrame) – PS Dataframe with LST_DIAG_CD column

  • lst_diag_cd_col_name (str)

Returns:

  • dask.DataFrame

  • list of str – List of CDPS category column names added.

Examples

>>> # Requires a dask DataFrame with diagnosis code list and aid columns
>>> df, lst_cols = CdpsRxRiskAdjustment.add_cdps_category_cols(
...     df, 'LST_DIAG_CD')
classmethod add_mrx_cat_cols(df: DataFrame, lst_ndc_col_name: str = 'LST_NDC') → Tuple[DataFrame, List[str]][source]

Adds MRX category columns

Parameters:

  • df (dask.DataFrame)

  • lst_ndc_col_name (str, default='LST_NDC')

Returns:

  • dask.DataFrame

  • list of str – List of MRX category column names added.

Examples

>>> # Requires a dask DataFrame with an NDC code list column
>>> df, lst_cols = CdpsRxRiskAdjustment.add_mrx_cat_cols(
...     df, 'LST_NDC')
classmethod calculate_risk(df: DataFrame, lst_cov: List[str], score_col_name: str = 'risk') → DataFrame[source]

Calculates CDPS risk adjustment score. The returned dataframe has the following new columns:

  • risk - CDPS risk score, dtype: float64

Parameters:

  • df (dask.DataFrame)

  • lst_cov (list of str) – List of covariates that CDPS MRX model uses

  • score_col_name (str, default='risk')

Return type:

dask.DataFrame

Examples

>>> # Requires a dask DataFrame with covariate columns and aid column
>>> df = CdpsRxRiskAdjustment.calculate_risk(
...     df, lst_cov=['a_under1', 'a_1_4'], score_col_name='risk')
classmethod combine_cdps_mrx_hierarchies(df: DataFrame) → DataFrame[source]

Apply CDPS MRX hierarchical rollups

Parameters:

df (dask.DataFrame)

Return type:

dask.DataFrame

Examples

>>> # Requires a dask DataFrame with CDPS and MRX category columns
>>> df = CdpsRxRiskAdjustment.combine_cdps_mrx_hierarchies(df)
data_folder = '/home/runner/work/medicaid-utils/medicaid-utils/medicaid_utils/adapted_algorithms/py_cdpsmrx/data'
filename = 'cdps_rx_risk_adjustment.py'
package_folder = '/home/runner/work/medicaid-utils/medicaid-utils/medicaid_utils/adapted_algorithms/py_cdpsmrx'
medicaid_utils.adapted_algorithms.py_cdpsmrx.cdps_rx_risk_adjustment.cdps_rx_risk_adjust(df: DataFrame, lst_diag_col_name: str = 'LST_DIAG_CD', lst_ndc_col_name: str = 'LST_NDC', score_col_name: str = 'risk') → DataFrame[source]

Calculate CDPS MRX risk adjustment score. This function expects the input dataframe to be aggregated to patient level, with columns containing comma separated lists of diagnosis codes & NDC codes during the observed time period.

Parameters:

  • df (dask.DataFrame) – BENE level Dataframe with diagnosis code & ndc_code list columns

  • lst_diag_col_name (str, default='LST_DIAG_CD') – Name of diagnosis code list column.

  • lst_ndc_col_name (str, default='LST_NDC') – Name of NDC code list column

  • score_col_name (str, default='risk') – CDPS MRX score output column name

Return type:

dask.DataFrame

Examples

>>> # Requires a patient-level dask DataFrame with diagnosis and NDC columns
>>> df = cdps_rx_risk_adjust(df, 'LST_DIAG_CD', 'LST_NDC')