Table 2 Analytical Methods used in order from early to most recent year of publication
From: Analytical methods to assess the impacts of activity-based funding (ABF): a scoping review
Author(s) (Year) | Analytical methodology | Number of data points (frequency) | Comparison group | Level of analysis | Outcome measures | |
|---|---|---|---|---|---|---|
Pre | Post | |||||
Shmueli et al. (2002) | ITS: OLS Regression | 2 (years) | 1 (year) | No | Hospital | LOS; case numbers; mortality; readmission; real hospital income |
Dismuke et al. (2002) | Count data models: Poisson & Negative Binomial regression | N/A | N/A | Yes: comparison of estimates between estimated models | Hospital | Mortality |
Theurl (2007) | ITS: Regression models accounting for hospital, regional & DRG-specific effects | 8 (years) | 6 (years) | No | Hospital | LOS |
Perelman et al. (2007) | ITS: Panel data regressions with hospital fixed effects | 4 (years) | 4 (years) | No | Hospital | LOS; medical cost; surgical cost |
Farrar et al. (2009) | DiD: Accounting for DRG & hospital trust fixed effects | 2/1 (years) | 1 (year) | Yes: English & Scottish hospitals not implementing PbR | Hospital | LOS; case numbers; mortality; readmission |
Martinussen et al. (2009) | ITS: OLS regression with hospital fixed effects | 3 (years) | 4 (years) | No | Procedure (DRG) | LOS |
Cooper et al. (2011) | DiD: OLS regression & IV estimation with hospital fixed effects | 4 (years) | 3 (years) | Yes: DRGs not affected by PbR | Patient (DRG) | Mortality |
Hamada et al. (2012) | DiD: Multilevel mixed-effect regression models | 3 (years) | 6 (years) | Yes: DRGs under FFS | Patient (DRG) | LOS; mortality; readmission; medical cost |
Yin et al. (2013) | Log-linear regression & pseudo-panel IV estimation | N/A | N/A | Yes: effects compared across three ischemic heart diseases | Patient (DRG) | LOS |
Kim H et al. (2015) | Descriptive methods: Descriptive statistics/ Student t-test/ Pearson chi-square test/ Fisher exact test | 6 (months) | 6 (months) | Yes: DRG effects before policy | Procedure (DRG) | LOS; mortality; readmission; outpatient visits; medical cost |
Kim J et al. (2015) | Descriptive methods: Descriptive statistics/ Student’s t-test/ Pearson chi-squared test | 8 (months) | 8 (months) | Yes: DRG effects before policy | Procedure (DRG) | LOS; readmission; hospital cost; outpatient visits; adverse events; complication rates |
Allen et al. (2016) | DiD: Differential-spline analyses to test for changes in the trend of an outcome | 25 (months) | 12 (months) | Yes: group of non-incentivised day surgeries | Procedure (DRG) | Case numbers |
Januleviciute et al. (2016) | ITS: Fixed-effects regression models | 2003 (base) | Â | No | Procedure (DRG) | Case numbers |
Sutherland et al. (2016) | ITS: Segmented GLM regression models accounting for Health Authority and hospital correlation over time | 26 (months) | 39 (months) | No | Hospital | LOS; case numbers; mortality; readmission |
Verzulli et al. (2017) | DiD: Regression models accounting for differences between short & medium-run hospital effects | 2 (years) | 4 (years) | Yes: DRGs not affected by price change | Hospital | LOS; case numbers |
Krabbe-Alkemade et al. (2017) | DiD: Log-linear regression models controlling for hospital market concentration & fixed effects | 2 (years) | 1 (year) | Yes: procedures under the budget reimbursement | Payment mechanism | Case number; DRG cost |
Jung et al. (2018) | Descriptive methods: Descriptive statistics/ Pearson chi- squared test/ Student’s t-test | 18 (months) | 6 (months) | No | Procedure | LOS; Readmission; |
Gaughan et al. (2019) | DiD & Synthetic Control approach | 4 (years) | 4.5 (years) | Yes: Non-incentivised same day discharge procedures | Procedure (DRG) | Case numbers |
Zeng (2019) | Logistic regression analysis & Descriptive methods | N/A | N/A | Yes: comparing profit & loss making DRGs | Patient (DRG) | LOS; DRG cost |