Table 1.
Changes in the US electronic health record market, 2006–10¶
| Market measures at the hospital referral region level | 2006 | 2007 | 2008 | 2009 | 2010 | Linear slope |
| Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | β (95% CI) | |
| All hospitals | ||||||
| Number of vendors | 3.70 (2.00) | 4.00 (2.10) | 4.07 (2.10)* | 4.21 (2.14)§ | 4.22 (2.17)§ | 0.12 (0.10 to 0.14)‡ |
| HHI** †† | 41.81 (21.56) | 39.82 (20.67) | 39.74 (20.65) | 38.37 (19.86)* | 38.28 (19.59)* | −1.02 (−1.30 to −0.75)‡ |
| % Hospitals using paper | 23.06 (21.47) | 15.21 (17.73)§ | 12.47 (15.73)§ | 8.28 (13.82)§ | 7.62 (12.68)§ | −3.78 (−4.13 to −3.43)‡ |
| Small hospitals (≤99 beds) | ||||||
| Number of vendors | 1.96 (1.40) | 2.16 (1.43) | 2.25 (1.48)* | 2.43 (1.58)§ | 2.48 (1.62)§ | 0.13 (0.11 to 0.15)‡ |
| HHI | 64.29 (29.12) | 62.46 (28.64) | 62.47 (29.14) | 59.74 (29.50)† | 59.23 (29.39)† | −1.94 (−2.40 to −1.48)‡ |
| % Hospitals using paper | 35.25 (33.97) | 26.13 (31.02)§ | 22.68 (29.40)§ | 16.97 (26.03)§ | 16.14 (25.38)§ | −4.74 (−5.44 to −4.05)‡ |
| Medium hospitals (100–249 beds) | ||||||
| Number of vendors | 2.23 (1.51) | 2.44 (1.57) | 2.44 (1.52)† | 2.59 (1.53)§ | 2.59 (1.57)§ | 0.09 (0.07 to 0.10)‡ |
| HHI | 62.05 (29.45) | 58.99 (1.54) | 58.63 (28.31) | 56.77 (28.26)* | 57.15 (28.44)* | −1.58 (−1.96 to −1.21)‡ |
| % Hospitals using paper | 18.70 (25.57) | 9.62 (18.79)§ | 7.77 (16.63)§ | 3.44 (10.89)§ | 2.98 (9.41)§ | −3.76 (−4.25 to −3.28)‡ |
| Large hospitals (≥ 250 beds) | ||||||
| Number of vendors | 2.16 (1.35) | 2.20 (1.45) | 2.21 (1.49) | 2.20 (1.43) | 2.22 (1.43) | 0.03 (0.01 to 0.04)‡ |
| HHI | 65.35 (29.83) | 65.08 (29.88) | 65.47 (30.18) | 64.99 (29.75) | 65.06 (29.63) | −0.44 (−0.71 to −0.18)‡ |
| % Hospitals using paper | 11.25 (24.45) | 6.28 (19.69)§ | 2.86 (12.89)§ | 0.66 (6.86)§ | 0.68 (6.92)§ | −2.68 (−3.14 to −2.21)‡ |
*
Statistically significantly different from 2006 (base year) at the 95% level.
†
Statistically significantly different from 2006 (base year) at the 90% level.
‡
Statistically significant slope at the 99% level.
§
Statistically significantly different from 2006 (base year) at the 99% level.
¶
Based on the vendors supplying data from a clinical repository that was live and operational, in the process of installation, or under contract centralized data repositories to hospitals.
**
Herfindahl–Hirschman Index.
††
Paper records were excluded in the HHI calculation.
HHI, Herfindahl–Hirschman Index.