Skip to main content

Table 2 Total number of units reporting with data, required for reconfirm after screening, with outliers corrected, and false alarm rate by different approach

From: Input data quality control for NDNQI national comparative statistics and quarterly reports: a contrast of three robust scale estimators for multiple outlier detection

RN hours per patient day by unit type IQR MAD FAST-MCD
  N 0 N 1 N 2 N 3 Post N 2 N 3 Post N 2 N 3 Post
Critical Care 1940 55 110 8 5.45% 109 8 5.30% 158 11 8.81%
Step Down 1259 22 60 2 4.69% 54 1 4.36% 87 2 7.37%
Other 4895 119 189 18 3.78% 179 17 3.68% 246 21 5.47%
Rehabilitation 451 8 18 0 3.99% 16 0 3.55% 16 0 3.95%
Neonatal 366 11 38 5 10.1% 38 5 10.1% 44 5 13.3%
Pediatric Critical Care 152 5 7 1 5.26% 7 1 5.26% 7 1 5.88%
Pediatric Step Down 33 1 6 0 18.2% 6 0 18.2% 6 0 25.8%
Pediatric Medical 99 3 5 0 5.05% 10 1 9.09% 19 1 20.7%
Pediatric Surgical 37 2 3 1 5.41% 3 1 5.41% 4 1 8.33%
Psychology ChildAd 373 11 24 2 7.69% 22 2 7.69% 26 1 9.91%
Psychology Gerip 117 4 10 1 5.15% 10 1 5.15% 11 1 10.1%
Falls Indicators
Fall Rate 8555 25 290 1 3.42% 286 1 3.37% 479 2 6.18%
Injury Fall Rate 8555 10 300 0 3.50% 397 0 4.62% 2249 5 28.9%
Fall Prior Risk Assmnt 8555 11 1039 7 12.1% - -   - -  
  1. Note: N0: Total number of units with data.
  2. N1: Total number of units with indicator value changed after data cleaning.
  3. N2: Total number of units identified by each method for potential outlier check.
  4. N3: Total number of units with indicator value checked and corrected (or dropped).
  5. *: Denote the method failed.
  6. Post: False alarm rates for post clean data.