Graduate Level Health Economics Econometric Solutions to Woolridge Computer Problems C6.4 C6.10 9.3 9.4 C9.1 C9.8

. use "E:\data\gpa2-1.dta"

. regress sat hsize hsizesq

      Source |       SS       df       MS              Number of obs =    4137

-------------+------------------------------           F(  2,  4134) =   15.93

       Model |  614822.097     2  307411.048           Prob > F      =  0.0000

    Residual |  79759024.2  4134  19293.4263           R-squared     =  0.0076

-------------+------------------------------           Adj R-squared =  0.0072

       Total |  80373846.3  4136  19432.7481           Root MSE      =   138.9

------------------------------------------------------------------------------

         sat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       hsize |   19.81446   3.990666     4.97   0.000     11.99061    27.63831

     hsizesq |  -2.130606    .549004    -3.88   0.000    -3.206949   -1.054263

       _cons |   997.9805   6.203448   160.88   0.000     985.8184    1010.143

------------------------------------------------------------------------------

. * Yes the quadratic term is statistically significant

. *The optimal high school size

. gen logsat = ln(sat)

. regress logsat hsize hsizesq

      Source |       SS       df       MS              Number of obs =    4137

-------------+------------------------------           F(  2,  4134) =   16.19

       Model |  .614405203     2  .307202602           Prob > F      =  0.0000

    Residual |  78.4287724  4134  .018971643           R-squared     =  0.0078

-------------+------------------------------           Adj R-squared =  0.0073

       Total |  79.0431776  4136   .01911102           Root MSE      =  .13774

------------------------------------------------------------------------------

      logsat |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       hsize |   .0196029   .0039572     4.95   0.000     .0118445    .0273612

     hsizesq |  -.0020872   .0005444    -3.83   0.000    -.0031546   -.0010199

       _cons |   6.896029   .0061515  1121.03   0.000     6.883969    6.908089

------------------------------------------------------------------------------

. *using the logsat as the DV the optimal high school size is 469

. clear

. use "E:\data\bwght2-1.dta"

. regress lbwght lbwght npvis npvissq

      Source |       SS       df       MS              Number of obs =    1764

-------------+------------------------------           F(  3,  1760) =       .

       Model |  74.2054098     3  24.7351366           Prob > F      =       .

    Residual |           0  1760           0           R-squared     =  1.0000

-------------+------------------------------           Adj R-squared =  1.0000

       Total |  74.2054098  1763   .04209042           Root MSE      =       0

------------------------------------------------------------------------------

      lbwght |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

      lbwght |          1          .        .       .            .           .

       npvis |  -7.94e-19          .        .       .            .           .

     npvissq |  -2.30e-20          .        .       .            .           .

       _cons |  -1.78e-15          .        .       .            .           .

------------------------------------------------------------------------------

. regress lbwght npvis npvissq

      Source |       SS       df       MS              Number of obs =    1764

-------------+------------------------------           F(  2,  1761) =   19.12

       Model |   1.5771321     2  .788566048           Prob > F      =  0.0000

    Residual |  72.6282777  1761  .041242634           R-squared     =  0.0213

-------------+------------------------------           Adj R-squared =  0.0201

       Total |  74.2054098  1763   .04209042           Root MSE      =  .20308

------------------------------------------------------------------------------

      lbwght |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       npvis |   .0189167   .0036806     5.14   0.000     .0116979    .0261355

     npvissq |  -.0004288     .00012    -3.57   0.000    -.0006641   -.0001934

       _cons |   7.957883   .0273125   291.36   0.000     7.904314    8.011451

------------------------------------------------------------------------------

. *Yes the quadratic term is statistically significant

. tab nvpis

variable nvpis not found

r(111);

. tab npvis

      total |

  number of |

   prenatal |

     visits |      Freq.     Percent        Cum.

------------+-----------------------------------

          0 |          5        0.28        0.28

          1 |          2        0.11        0.40

          2 |          5        0.28        0.68

          3 |         12        0.68        1.36

          4 |          6        0.34        1.70

          5 |         27        1.53        3.23

          6 |         59        3.34        6.58

          7 |         58        3.29        9.86

          8 |        117        6.63       16.50

          9 |         96        5.44       21.94

         10 |        199       11.28       33.22

         11 |        115        6.52       39.74

         12 |        618       35.03       74.77

         13 |         72        4.08       78.85

         14 |         97        5.50       84.35

         15 |        143        8.11       92.46

         16 |         41        2.32       94.78

         17 |         12        0.68       95.46

         18 |         15        0.85       96.32

         19 |          4        0.23       96.54

         20 |         35        1.98       98.53

         21 |          5        0.28       98.81

         22 |          2        0.11       98.92

         23 |          1        0.06       98.98

         24 |          2        0.11       99.09

         25 |          3        0.17       99.26

         26 |          1        0.06       99.32

         30 |          7        0.40       99.72

         33 |          1        0.06       99.77

         35 |          1        0.06       99.83

         36 |          1        0.06       99.89

         40 |          2        0.11      100.00

------------+-----------------------------------

      Total |      1,764      100.00

. gen visits = .

(1832 missing values generated)

. replace visits = 1 if npvis >= 22

(89 real changes made)

. tab visits

     visits |      Freq.     Percent        Cum.

------------+-----------------------------------

          1 |         89      100.00      100.00

------------+-----------------------------------

      Total |         89      100.00

. tab npvis if npvis >21

      total |

  number of |

   prenatal |

     visits |      Freq.     Percent        Cum.

------------+-----------------------------------

         22 |          2        9.52        9.52

         23 |          1        4.76       14.29

         24 |          2        9.52       23.81

         25 |          3       14.29       38.10

         26 |          1        4.76       42.86

         30 |          7       33.33       76.19

         33 |          1        4.76       80.95

         35 |          1        4.76       85.71

         36 |          1        4.76       90.48

         40 |          2        9.52      100.00

------------+-----------------------------------

      Total |         21      100.00

. tab npvis visits

     total |

 number of |

  prenatal |   visits

    visits |         1 |     Total

-----------+-----------+----------

        22 |         2 |         2

        23 |         1 |         1

        24 |         2 |         2

        25 |         3 |         3

        26 |         1 |         1

        30 |         7 |         7

        33 |         1 |         1

        35 |         1 |         1

        36 |         1 |         1

        40 |         2 |         2

-----------+-----------+----------

     Total |        21 |        21

. *Yes it does make sense in that if there are thatmany visits it could be a problemat

> ic pregnancy with the likelihood of lower birth outcomes

. regress lbwght npvis npvissq mage magesq

      Source |       SS       df       MS              Number of obs =    1764

-------------+------------------------------           F(  4,  1759) =   11.56

       Model |  1.90136387     4  .475340968           Prob > F      =  0.0000

    Residual |  72.3040459  1759    .0411052           R-squared     =  0.0256

-------------+------------------------------           Adj R-squared =  0.0234

       Total |  74.2054098  1763   .04209042           Root MSE      =  .20274

------------------------------------------------------------------------------

      lbwght |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       npvis |   .0180374   .0037086     4.86   0.000     .0107636    .0253112

     npvissq |  -.0004079   .0001204    -3.39   0.001    -.0006441   -.0001717

        mage |    .025392   .0092542     2.74   0.006     .0072417    .0435423

      magesq |  -.0004119   .0001548    -2.66   0.008    -.0007154   -.0001083

       _cons |   7.583713   .1370568    55.33   0.000     7.314901    7.852524

------------------------------------------------------------------------------

. *The optimal age is 31

. gen age = .

(1832 missing values generated)

. replace age = 1 if mage > 31

(605 real changes made)

. replace age = 0 if mage <= 31

(1227 real changes made)

. tab age

        age |      Freq.     Percent        Cum.

------------+-----------------------------------

          0 |      1,227       66.98       66.98

          1 |        605       33.02      100.00

------------+-----------------------------------

      Total |      1,832      100.00

. regress bwght npvis npvissq mage magesq

      Source |       SS       df       MS              Number of obs =    1764

-------------+------------------------------           F(  4,  1759) =    8.59

       Model |  11376019.1     4  2844004.78           Prob > F      =  0.0000

    Residual |   582383777  1759  331087.992           R-squared     =  0.0192

-------------+------------------------------           Adj R-squared =  0.0169

       Total |   593759796  1763  336789.448           Root MSE      =   575.4

------------------------------------------------------------------------------

       bwght |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       npvis |   37.47386   10.52531     3.56   0.000     16.83042     58.1173

     npvissq |  -.7862864   .3417884    -2.30   0.022     -1.45664   -.1159322

        mage |   81.60554   26.26395     3.11   0.002     30.09369    133.1174

      magesq |  -1.327179   .4392829    -3.02   0.003     -2.18875   -.4656073

       _cons |   1860.381    388.977     4.78   0.000     1097.475    2623.286

------------------------------------------------------------------------------

9.3

       Let math10 denote the percentage of students at a Michigan high school receiving a passing score on a standardized math test (see also Ex 4.2).  We are interested in estimating the effect of per student spending on math performance.  A simple model is

i)       You need to be “poor” to be in the federally funded student lunch program.

ii)      Because lnchprg is negatively correlated with log(expend).  It is significant.

iii)    Yes

iv)      Math10 would decrease by 3.24% if lunch program increase by 10%.

v)       It’s a good predictor variable because R-squared is increasing.

9.4

(i)

Tvhours= tvhours*+e0

(ii)

It’s not likely to hold in the example, because if the tvhours=0 then they would be reported as zero

So the error depends on the actual tvhours

C9.1

(i)

. use "C:\Users\sphl\Desktop\CEOSAL1-1.DTA"

. generate rosneg = 0

. replace rosneg = 1 if (ros<1)

(23 real changes made)

. regress lsalary lsales roe rosneg

      Source |       SS       df       MS              Number of obs =     209

-------------+------------------------------           F(  3,   205) =   28.81

       Model |  19.7902019     3  6.59673397           Prob > F      =  0.0000

    Residual |  46.9319613   205  .228936397           R-squared     =  0.2966

-------------+------------------------------           Adj R-squared =  0.2863

       Total |  66.7221632   208  .320779631           Root MSE      =  .47847

------------------------------------------------------------------------------

     lsalary |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

      lsales |   .2883868   .0336172     8.58   0.000      .222107    .3546665

         roe |   .0166571   .0039681     4.20   0.000     .0088336    .0244806

      rosneg |   -.225675    .109338    -2.06   0.040    -.4412462   -.0101038

       _cons |   4.297602   .2932526    14.65   0.000     3.719425     4.87578

------------------------------------------------------------------------------

. ovtest

Ramsey RESET test using powers of the fitted values of lsalary

       Ho:  model has no omitted variables

                 F(3, 202) =      1.07

                  Prob > F =      0.3614

** We accept the null-hypothesis

C9.8

   1.  New update available; type -update all-

. use "C:\Users\sphl\Desktop\twoyear.dta"

. codebook stotal

--------------------------------------------------------------------------------------

stotal                                                   total standardized test score

--------------------------------------------------------------------------------------

                  type:  numeric (float)

                 range:  [-3.3247969,2.2353656]       units:  1.000e-09

         unique values:  227                      missing .:  0/6763

                  mean:   .047483

              std. dev:   .853544

           percentiles:        10%       25%       50%       75%       90%

                          -1.10531  -.327343         0   .610791   1.13706

. reg stotal jc

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  1,  6761) =    1.03

       Model |  .752020193     1  .752020193           Prob > F      =  0.3097

    Residual |   4925.6191  6761  .728534108           R-squared     =  0.0002

-------------+------------------------------           Adj R-squared =  0.0000

       Total |  4926.37112  6762  .728537581           Root MSE      =  .85354

------------------------------------------------------------------------------

      stotal |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |    .013658   .0134431     1.02   0.310    -.0126946    .0400107

       _cons |   .0428543   .0113348     3.78   0.000     .0206344    .0650741

------------------------------------------------------------------------------

. reg stotal univ

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  1,  6761) = 1574.72

       Model |  930.653211     1  930.653211           Prob > F      =  0.0000

    Residual |  3995.71791  6761  .590995106           R-squared     =  0.1889

-------------+------------------------------           Adj R-squared =  0.1888

       Total |  4926.37112  6762  .728537581           Root MSE      =  .76876

------------------------------------------------------------------------------

      stotal |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

        univ |   .1615085     .00407    39.68   0.000       .15353    .1694869

       _cons |  -.2636267   .0122004   -21.61   0.000    -.2875434     -.23971

------------------------------------------------------------------------------

. reg lwage jc univ exper stotal

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  4,  6758) =  500.23

       Model |  367.406832     4  91.8517079           Prob > F      =  0.0000

    Residual |  1240.88926  6758  .183617825           R-squared     =  0.2284

-------------+------------------------------           Adj R-squared =  0.2280

       Total |  1608.29609  6762  .237843255           Root MSE      =  .42851

------------------------------------------------------------------------------

       lwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |   .0630514   .0068214     9.24   0.000     .0496792    .0764235

        univ |   .0686405   .0025651    26.76   0.000     .0636121     .073669

       exper |   .0048773   .0001571    31.04   0.000     .0045692    .0051854

      stotal |   .0493766   .0068096     7.25   0.000     .0360277    .0627255

       _cons |   1.495271   .0212176    70.47   0.000     1.453678    1.536864

------------------------------------------------------------------------------

. test jc + univ <0

+ not found

r(111);

. test jc+ univ<0

+ not found

r(111);

. test jc < univ

< not found

r(111);

. test jc + univ < 0

+ not found

r(111);

. reg lwage jc totcoll exper stotal

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  4,  6758) =  500.23

       Model |  367.406831     4  91.8517079           Prob > F      =  0.0000

    Residual |  1240.88926  6758  .183617825           R-squared     =  0.2284

-------------+------------------------------           Adj R-squared =  0.2280

       Total |  1608.29609  6762  .237843255           Root MSE      =  .42851

------------------------------------------------------------------------------

       lwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |  -.0055892   .0069385    -0.81   0.421    -.0191909    .0080126

     totcoll |   .0686405   .0025651    26.76   0.000     .0636121     .073669

       exper |   .0048773   .0001571    31.04   0.000     .0045692    .0051854

      stotal |   .0493766   .0068096     7.25   0.000     .0360277    .0627255

       _cons |   1.495271   .0212176    70.47   0.000     1.453678    1.536864

------------------------------------------------------------------------------

. gen stotalsq = stotal * stotal

. reg lwage jc univ exper stotal stotalsq

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  5,  6757) =  400.17

       Model |  367.436832     5  73.4873664           Prob > F      =  0.0000

    Residual |  1240.85926  6757  .183640559           R-squared     =  0.2285

-------------+------------------------------           Adj R-squared =  0.2279

       Total |  1608.29609  6762  .237843255           Root MSE      =  .42853

------------------------------------------------------------------------------

       lwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |   .0632092    .006833     9.25   0.000     .0498143    .0766041

        univ |   .0685131   .0025846    26.51   0.000     .0634466    .0735797

       exper |   .0048776   .0001572    31.04   0.000     .0045695    .0051857

      stotal |   .0501562   .0070778     7.09   0.000     .0362814     .064031

    stotalsq |   .0019191   .0047481     0.40   0.686    -.0073886    .0112268

       _cons |    1.49399   .0214545    69.64   0.000     1.451932    1.536047

------------------------------------------------------------------------------

. gen jcstotal = jc*stotal

. gen univtotal = univ*stotal

. reg lwage jc univ exper stotal jctotal univtotal

variable jctotal not found

r(111);

. reg lwage jc univ exper stotal jcstotal univtotal

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  6,  6756) =  334.24

       Model |  368.126196     6  61.3543661           Prob > F      =  0.0000

    Residual |   1240.1699  6756  .183565704           R-squared     =  0.2289

-------------+------------------------------           Adj R-squared =  0.2282

       Total |  1608.29609  6762  .237843255           Root MSE      =  .42845

------------------------------------------------------------------------------

       lwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |    .063636   .0068983     9.22   0.000     .0501132    .0771588

        univ |   .0689706   .0027439    25.14   0.000     .0635917    .0743494

       exper |   .0048856   .0001572    31.08   0.000     .0045775    .0051938

      stotal |   .0582108   .0087001     6.69   0.000     .0411559    .0752657

    jcstotal |  -.0168818    .009275    -1.82   0.069    -.0350638    .0013001

   univtotal |  -.0026652   .0029894    -0.89   0.373    -.0085254     .003195

       _cons |   1.495924   .0212432    70.42   0.000      1.45428    1.537567

------------------------------------------------------------------------------

. test jcstotal univtotal

 ( 1)  jcstotal = 0

 ( 2)  univtotal = 0

       F(  2,  6756) =    1.96

            Prob > F =    0.1410

. * not jointly significant

. reg lwage jc univ exper stotal

      Source |       SS       df       MS              Number of obs =    6763

-------------+------------------------------           F(  4,  6758) =  500.23

       Model |  367.406832     4  91.8517079           Prob > F      =  0.0000

    Residual |  1240.88926  6758  .183617825           R-squared     =  0.2284

-------------+------------------------------           Adj R-squared =  0.2280

       Total |  1608.29609  6762  .237843255           Root MSE      =  .42851

------------------------------------------------------------------------------

       lwage |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

          jc |   .0630514   .0068214     9.24   0.000     .0496792    .0764235

        univ |   .0686405   .0025651    26.76   0.000     .0636121     .073669

       exper |   .0048773   .0001571    31.04   0.000     .0045692    .0051854

      stotal |   .0493766   .0068096     7.25   0.000     .0360277    .0627255

       _cons |   1.495271   .0212176    70.47   0.000     1.453678    1.536864

------------------------------------------------------------------------------

. * the original regression is the preferred one

.

3 Ways to Be a Better Writer

Do you wish you were a better writer, but always feel like something is holding you back? Chances are, whatever is keeping you from writing the novel, blog or thesis that you wish you could write are the common factors that all writers face: physiological needs, thought patterns that block creativity, or self-defeating habits that you’ve developed. Let’s talk about how these can be overcome.

1. Body Needs: Identify Them,  Satisfy Them, Write

What holds me back from writing in terms of antecedent behavior to writing are a million physiological needs. I’m hungry, I have to go to the bathroom, my hip or back, or neck hurt. All of these have a solution. Why fight them? Fix them.

2. Too Many (Or Not Enough) Ideas: Back of the Napkin Method

Whether it is too many or not enough ideas, the way to overcome this problem is simple. If you have  too many ideas, write them down. All of them. One by one. In a circle, in a list, in little thought clouds. If you don’t have enough ideas, start drawing, or just write any words that cross your mind.  Unless you are under the influence of mind altering substances, there are always things floating around in there. Set a timer for 10 minutes and sit with a blank piece of paper or black screen. Don’t self-censor. Feel the fear and do it anyway.

Courtesy Flickr Melo McC

3. It Takes Ten Minutes to Get Coffee at Starbucks

Once you have taken care of 1 and 2 above, think about ways you can work this into your daily routine. Most humans are on autopilot at least 50% of the time anyway.  Make it the first thing you do at work when you come in, or make it the last thing you do before you leave. Your job isn’t going to vaporize into the void in 10 minutes. Emails can wait. Whatever your writing goals, you already have the keys inside yourself to overcome them.  The way you unlock that potential is to set yourself up for success. Whatever the behavioral cue that helps you get it done is ok. Give yourself permission, go ahead and write it out.

5 Inconvinient Truths No One Tells You About Being a Mom

Sometimes I look at the happy pictures of me with my son and can't even really believe it is me. That it is us. Somewhere along the line, the industrial-parent complex forgot to tell me the truth about being a mom. So that you won't have to endure the same rediculosity* in silence, here is my top 5 list that doesn't get talked about:

1) Being a mom is hard work. It doesn't matter how great your partner is, there are certain things only the female can do. So unless you are in a lesbian couple and your partner is having induced lactation, only the person who gives birth can breastfeed. Some times the screams of "mom, mom, mom" will only quiet to the mom. Being a mom is a 24 hour a day, 7 days a week activity. Even when you are not with your child you may feel anxious about that. So there is no winning. You will do lots of laundry, sacrifice your schedule and your body, even vacations are work.

2) Forget sleeping. Maybe you've tried crying it out and that worked for you. Maybe your child is an angel that peacefully slumbers 14 hours a day. Good for you. No one warns you that you could possibly go more than 18 months without a good night's sleep.

3) Priorities will change. You can't really understand this until you are a mom. But priorities will change. Trust me.

4) You may feel lonely and angry. Even if you go to parent groups and play dates, becoming a family is unique to each of the members in the family. You might be in a same-sex relationship or a single mom or your spouse might work a lot. Delivering childcare can be lonely by yourself. If you are an introvert, attending play dates might be exhausting. Pretty much either way it is not fun. Your play date mates might experience family in a different way that makes the feel more connected and happy than you; that might make you feel jealous! Just remember, you are NOT bad for feeling sad or  lonely or angry. What matters is how you deal with those feelings. There is no easy answer to your dark feelings. You must live with them everyday and try to work through them.

5) There is no undo. While technically you can drop them off at a local hospital if it is just too much, there is no magic button you can push to make your life go back to pre-parenthood. Once you hold them in your arms, you will be changed forever.

*Bush-ism, a word invented by George W. Bush

Christmas, 2014

Clinicians Perform “Chart Biopsies” to Prepare for Patient Handoffs

An AHRQ-funded study examines the practice of “chart biopsies,” a process that allows receiving clinicians to understand patients’ clinical course prior to handoff. According to the article, “Chart Biopsy: An Emerging Medical Practice Enabled by Electronic Health Records and Its Impacts on Emergency Department-Inpatient Admission Handoffs,” the term ”chart biopsy” refers to the activity of examining portions of a patient’s electronic health record (EHR) to gather specific information about that patient or to develop a broader understanding of the patient’s care. With many hospitals moving to EHRs, practitioners are now able to view patient records prior to the physical transfer of the patient. Chart biopsies enable receiving clinicians to gain a stronger understanding of the patient’s condition, allow receiving practitioners to better prepare for handoff and subsequent care, and guard against possible bias in verbal reports. Chart biopsy is an emerging practice that is not yet available as a tool in all hospitals. However, proponents point out that chart biopsy provides a tool to enrich coordination and collaboration, which may enable safety, efficiency, and effectiveness of medical care. To access the article, which appeared in the September 2012 issue of the Journal of the American Medical Informatics Association, select: http://1.usa.gov/PT9qJH.

Launch of Open Access Journal on Electronic Clinical Data

The Agency for Healthcare Research and Quality and the Electronic Data Methods Forum are pleased to announce the official launch of eGEMs (Generating Evidence and Methods to improve patient outcomes). eGEMs  is an open access journal focused on using electronic clinical data to advance research and quality improvement, with the overall goal of improving patient and community outcomes. Authors are welcome to submit papers, images, or other media focused on data methods, informatics, governance, and the learning health system.

Select to access: http://repository.academyhealth.org/egems/

Clinical Quality Measures Useful in Estimating Quality Across Different EHRs

A recent study supported by the Agency for Healthcare Research and Quality examined the sensitivity and specificity of clinical quality measures generated by electronic health records. “Accuracy of Electronically Reported “Meaningful Use” Clinical Quality Measures: A Cross-Sectional Study” appeared in the Annals of Internal Medicine on January 15. The study looked at the accuracy of interoperable electronic reporting for 12 quality measures, finding that 9 of the clinical quality measures showed good consistency in estimating quality of care. This suggests that electronic quality measurement has the capability to identify areas in need of improvement and quantify the impact of changes that have been made. This research was a part of AHRQ’s ongoing efforts to study ways to improve electronically enabled quality measurement. For a short video about this project, select: http://healthit.ahrq.gov/EQMKaushalVideo.

  

The abstract is available at: http://www.ncbi.nlm.nih.gov/pubmed/23318309.