Admission and Mid-Stay MedisGroupso Scores as Predictors of Death Within 30 Days of Hospital Admission

Lisa I. Iezzoni, MD, MS, Arlene S. Ash, PhD, Gerald Coffinn, MS, and MarkA. Moskowitz, MD

Introduction The Federal Government annually publishes mortality rates of almost 6,000 United States hospitals caring for Medicare beneficiaries. Although release of the mortality data is accompanied by numerous and appropriate disclaimers about their utility, the data are clearly intended to serve as an initial clue about the quality of hospital care, especially in an era of tightening cost constraints. The first hospital mortality statistics, published in March 1986, indicated that 142 hospitals had significantly higher death rates than predicted and 127 had significantly lower rates. ' However, the facility with the most aberrant death rate (87.6 percent of Medicare patients dying compared with a predicted 22.5 percent) was a hospice caring for terminally ill patients. The government model had not adequately considered severity of illness. This incident prompted a flurry of research activity, which has resulted in more sophisticated and methodologically sound approaches for producing the Medicare hospital mortality statistics.2 Since 1987, the model has adjusted for patient age, secondary diagnoses, and a variety of other characteristics available in Medicare's administrative files; in addition, to control better for case mix and to accommodate the possibility that hospitals may perform better in some clinical areas than in others, condition-specific hospital mortality rates are also produced for selected diseases. However, data limitations remain. The information contained in Medicare's administrative files is confined to basic demographic, diagnostic, and procedural data not considered adequate to the task of severity adjustment by some critics.3-7 The purpose of this research

was to examine the performance of MedisGroups®, a severity measurement system based on data drawn from the medical record, as a predictor of mortality.

Methods Description of MedisGroups MedisGroups assesses severity as the risk of imminent organ failure independent of diagnosis, using information on key clinical findings (KCFs) abstracted from the medical record. MedisGroups is a registered trade name; the system is owned and marketed by MediQual Systems, Inc., in Westborough, Massachusetts. Severity is scored on admission on a five-point scale ranging from 0 to 4. For patients who remain hospitalized for the required period, a second severity rating is performed, the so-called "mid-stay" review, generating an appraisal of extent of "morbidity" (no morbidity, morbid, or major morbidity). MedisGroups is described in detail elsewhere,8-" and more information on the system is presented in an appendix. This description and the data used for this study pertain to the 1987 version of MedisGroups. Address reprint requests to Lisa I. Iezzoni, Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Hospital, 330 Brookline Avenue, Boston, MA 02215, Tel: (617)735-4476. Dr. Iezzoni is also responsible for negotiations concerning this manuscript. All the authors were with the Health Care Research Unit, Section of General Internal Medicine, Evans Memorial Department of Clinical Research and Medicine, Boston University School of Medicine. Dr. lezzoni was also with the Health Policy Institute, Boston University School of Public Health. This paper, submitted to the Journal January 10, 1990, was revised and accepted for publication June 12, 1990.

January 1991, Vol. 81, No. 1

MedisGroups Scores as Predictors of Death

When MedisGroups is used for quality assessment, one strategy is to contrast the admission and mid-stay scores to determine whether the patient worsened, improved, or remained the same. For this purpose, admission scores of 0 and 1 are equated with no morbidity on the midstay review, while a 2 corresponds to morbid, and 3 and 4 correspond to major morbidity. In our analyses, we examined the predictive power of the admission MedisGroups score alone, the mid-stay review score alone, and the admission and mid-stay review scores considered together.

Descnption of the Data Base The data base used for this study was created through the Health Care Financing Administration (HCFA) Health Standards and Quality Bureau. The data were collected by the peer review organizations (PROs) in Alabama, Arizona, Indiana, New York, Pennsylvania, Utah, and Wisconsin; the states were chosen to provide a geographically representative sample.

Cases were randomly selected using administrative data on all Medicare beneficiaries, including disabled individuals younger than 65 years of age, admitted between January 1985 and May 1986. Selection was predicated on the presence of specified diagnostic and procedural codes. Final case selection was dictated by the PROs based on such factors as the availability of medical records. The analytic data base included 20,985 cases from 833 acute care hospitals across the seven states. The number of cases per hospital ranged from one to 480, with a median of 13 cases (mean = 25, standard deviation = 39) per hospital. Conditions were selected based on their volume, relevance to Medicare ben-

January 1991, Vol. 81, No. 1

eficiaries (such as the elderly), and complexity of clinical issues raised in their care. The six study conditions were pneumonia or chronic obstructive pulmonary disease (COPD); coronary artery revascularization (including coronary artery bypass graft surgery and percutaneous transluminal coronary angioplasty); acute myocardial infarction; congestive heart failure; cholecystectomy; and prostate surgery (including transurethral and open prostatectomies). Cases in three conditions (pneumonia/COPD, acute myocardial infarction, and congestive heart failure) were selected using the principal diagnosis, while selection in the remaining conditions was based on procedure codes. Because of this selection approach, it was possible for cases chosen by diagnosis codes to have had a major procedure as well. MedisGroups information and other selected clinical variables were collected by PRO abstractors trained by MediQual Systems, Inc. MediQual requires that reviewers achieve 95 percent- exact agreement for items abstracted from the medical record; however, we did not independently confirm the reliability of the data. Data entrywas performed by the PROs, then transmitted to MediQual Systems, Inc., for creation of a data tape for HCFA. The tape included admission and mid-stay scores, as well as the values for all KCFs.

Definition of Mortality The data included date of death information from the Social Security administrative files, making it possible to identify deaths that occurred following discharge. The analyses focused on death within 30 days of admission, regardless of whether the death occurred in hospital.

The 30-day figure is currently used by HCFA in producing its annual hospital mortality statistics, and it mninimizes concerns about how differences in discharge practices across institutions and regions influence in-hospital death information.2

Statistical Methods All analyses were performed within each of the six conditions. Each of the three models used the MedisGroups scores as categories indicated by dummy variables: admission score, five categories (0, 1, 2, 3, and 4); mid-stay review score, four categories (no mid-stay review, no morbidity, morbid, and major morbidity); and the joint categorization by both the admission and mid-stay scores, 20 categories (the five admission score classes crossed with the four mid-stay score classes). Given that each of our models contained only dummy variables as predictors, ordinary least squares and logistic regression methods produced exactly the same predicted probability of dying for every case (i.e., the same fraction dying among all cases in the same category). We looked at both the R-squared values and the C statistic from a logistic regression12,13 to examine the ability of the three models based on MedisGroups scores to explain differences in death within 30 days of admission. However, only the R-squared values are reported here since no substantive differences emerged from looking at the C statistic.10

Results Death Rates across MedisGroups Scores The fraction of patients dying within 30 days of admission varied by condition,

American Journal of Public Health 75

Iezzoni, et al.

with the surgical conditions having lower death rates (1.1-5.8 percent) than the medical conditions (12.8-23.8 percent), (Table 1). The distribution of cases across admission severity scores also varied by condition. For example, only 6.1 percent of the surgical cases had high admission scores of 3 or 4, compared with 30.4 percent of the medical cases. Within all six conditions, higher admission scores were generally associated with higher death rates. However, even among cases with identical admission scores, those in conditions with higher overall death rates generally had higher severity-score-specific death rates. For example, among the six conditions, prostate surgery cases had the lowest overall death rate (1.1 percent) while acute myocardial infarction cases had the highest (23.8 percent). Not only did prostatectomy contain far fewer cases with scores of 3 and 4 than did acute myocardial infarction (1.6 percent of cases compared with 33.2 percent), but also, within each of the five admission scores, the rate of dying was less than one-fifth as great for prostatectomy cases as for infarction cases. The rates ofdeath by mid-stay review score also varied widely across conditions (Table 2), as did the distribution of cases across mid-stay scores. Even with major morbidity, patients in the three surgical conditions were much less likely to die than their medical counterparts. Death rates rose somewhat between no morbidity and morbid, but the rates increased at least threefold between morbid and major morbidity. Except in cholecystectomy and prostate surgery cases, many of the patients with no mid-stay review died. For instance, 72.8 percent of acute myocardial

76 American Journal of Public Health

infarction patients without a mid-stay review died, as did 31.4 percent of coronary revascularization patients. Thus, the reason for the absence of a mid-stay review was often the early death of the patient. Nevertheless, most of the patients who died remained alive long enough to have a mid-stay review. In three conditions, the plurality of patients who died had major morbidity. In cholecystectomy and prostate surgery patients, the 1 percent death rate among the many cases with no morbidity represented more deaths than the over 14 percent death rates among the few cases with major morbidity. In contrast, 33.6 percent of acute myocardial infarction deaths occurred among patients with major morbidity, while 38.1 percent of deaths occurred among patients with no midstay review.

&aplaining Variations in Rates of Death When admission scores were used to predict 30-day mortality (Table 3), the R-squared values ranged from 0.01 in cholecystectomy and prostate surgery cases to 0.16 in acute myocardial infarction cases. In comparison, the mid-stay review

score was a more powerful predictor ofrisk of death, with R-squared values ranging from a low of 0.03 for prostate surgery to a high of 0.34 for acute myocardial infarction. Combining information about the admission and mid-stay scores generally yielded much greater predictive power than did either score alone, producing the highest R-squared value in all conditions. Given the large fraction of deaths among those without a mid-stay review, we looked at whether the fact of a midstay review alone was predictive of death. We regressed 30-day mortality on a dichotomous variable indicating the presence or absence of a mid-stay review. For acute myocardial infarction cases, the R-squared value produced was 0.19. For other conditions, however, the R-squared value was much smaller: 0.04 for pneumonia/COPD and coronary revascularization, 0.02 for congestive heart failure, and 0.00 for cholecystectomy and prostate surgery. Thus, in all conditions, the simple fact of a mid-stay review did not appear to cause the greater predictive power of the mid-stay review cat-

egories. January 1991, Vol. 81, No. 1

MedisGroups Scores as Predictors of Death

Discussion The superior performance of the two approaches involving the mid-stay review was not primarily related to the fact that the absence of a mid-stay review identified patients who died quickly, except among acute myocardial infarction patients. Probably more important is the reliance of the mid-stay review score on physiological derangements generally one week into the hospitalization. It seems almost a clinical truism that patients with severe problems this far into the admission would have worse prognoses than patients without such findings. By contrast, however, admission scores include severe physiological derangements that can often be treated effectively with current medical therapeutics. This conjecture is consistent with the relatively weak performance of the admission MedisGroups score as a predictor of mortality. In general, patients in the three surgical conditions were much less likely to die, despite high admission scores. The reason may be that those acute physiological derangements that present on admission (and are reflected in high admission scores) represent manifestations of the patients' underlying disease, which will be explicitly addressed by the surgery, usually with a successful outcome (especially for cholecystectomy and prostate surgery). In contrast, certain physiological derangements stemming from acute and chronic medical ailments may be less amenable to control and resolution. This study has several limitations. First, the number of deaths was relatively small, especially for the surgical conditions. Thus, the R-squared values reported here (particularly for the combined admission and mid-stay score model) may be inflated by overfitting. Second, the results cannot be generalized to younger patients, although the case selection strategy included a small number of Medicare beneficiaries younger than 65 years of age. Finally, MedisGroups was developed to reflect events during a hospitalization. By focusing on death within 30 days of admission, we included some events that occurred following discharge. When analyses were repeated using in-hospital death, admission scores and mid-stay scores produced R-squared values virtually identical to those for 30day mortality, except for acute myocardial infarction cases, for which the R-squared value for the mid-stay score

January 1991, Vol. 81, No. 1

was 0.39 for in-hospital death compared with 0.34 for 30-day mortality. Similarly, for the combination of admission and mid-stay scores the R-squared values were only 0.01 to 0.03 higher for predictions of in-hospital mortality, except in acute myocardial infarction cases, for which the R-squared value was 0.47 for in-hospital death compared with 0.41 for 30-day mortality. Even if the mid-stay review score were perfectly predictive of mortality, it is not at all clear that it would be an appropriate risk adjuster for widespread quality assessment efforts, such as the federal hospital-level mortality figures. The mid-stay review rating could be affected by shortfalls in the quality of care, the very event that mortality statistics seek to identify. Therefore, adjusting mortality figures by the mid-stay review score could mask instances of quality problems. For such risk adjustment, the appropriate construct is the admission "severity" score, which performs relatively poorly as a predictor of 30-day mortality in this data base. These findings could be contrasted with those produced by other models of admission severity, such as the Medicare Mortality Predictor System14"15 and the Acute Physiology and Chronic Health Evaluation,16-8 but such comparisons would be hampered by differences in the data bases, predictive models, and statistical techniques used in the different studies. No published research yet exists comparing the performance of various severity systems as predictors of 30-day mortality on comparable data bases across conditions. Although these results must be viewed as preliminary and pertinent only to Medicare beneficiaries, they may have important implications for states and regions that have mandated the collection of severity information for assessing hospital quality, especially if one purpose is to examine severity-adjusted mortality rates. For example, Pennsylvania and Iowa now require hospitals to collect patient severity information using MedisGroups, with one goal being to examine severity-adjusted outcomes. Our findings raise the crucial question: What causes the large amount of variation in death rates unexplained by the MedisGroups admission score? Further study is needed to answer this question so that the MedisGroups information can be used in an appropriate and valid fashion. [1

Acknowledgments We are grateful to Henry Krakauer, MD, PhD, Health Standards and Quality Bureau, Health Care Financing Administration, for furnishing the data base and assisting us in its use. We also thank Elizabeth Estabrook, RN, and Paul Steen, MD, of MediQual Systems, Inc., for their assistance and comments. This research was supported by the Health Care Financing Administration, Office of Research, under Cooperative Agreement No. 18-C-98526/1-05. The views expressed are solely those of the authors.

References 1. Brinkley J: U.S. releasing lists of hospitals with abnormal mortality rates. New York Times March 12, 1986. 2. Dubois RW: Hospital mortality as an indicator of quality. In: Goldfield N, Nash DB (eds): Providing Quality Care: The Challenge to Clinicians. Philadelphia: American College of Physicians, 1989; 107-132. 3. Mullin RL: Diagnosis-related groups and severity. ICD-9-CM, the real problem. JAMA 1985; 254:1208-1210. 4. McMahon LF, Smits HL: Can Medicare prospective payment survive the ICD9-CM disease classification system? Ann Intern Med 1986; 104:562-566. 5. Gertman PM, Lowenstein S: A research paradigm for severity of illness: Issues for the diagnosis-related group system. Health Care Fin Rev 1984; (ann suppl) 79-90. 6. Feinstein AR: ICD, POR, and DRG. Unsolved scientific problems in the nosology of clinical medicine. Arch Intern Med 1988; 148:2269-2274. 7. Green J, Wintfeld N, Sharkey P, Passman UJ: The importance of severity of illness in assessing hospital mortality. JAMA 1990; 263:241-246. 8. lezzoni LI, Moskowitz MA: A clinical assessment of MedisGroups. JAMA 1988; 260;3159-3163. 9. Iezzoni LI, Moskowitz MA, Ash AS: The ability of MedisGroups and its clinical variables to predict cost and in-hospital death. Report prepared for the Health Care Financing Administration, Cooperative Agreement No. 18-C-98526/1-04. Boston: Health Care Research Unit, Boston University Medical Center, 1988. 10. Iezzoni LI, Ash AS, Moskowitz MA: Admission and mid-stay MedisGroups scores as predictors of hospital charges and 30-day mortality. Report prepared for the Health Care Financing Administration, Cooperative Agreement No. 18-C-98526/1-05. Boston: Health Care Research Unit, Boston University Medical Center, 1989. 11. Brewster AC, Karlin BG, Hyde LA, Jacobs CM, Bradbury RC, Chae YM: MEDISGRPS: A clinically based approach to classifying hospital patients at admission. Inquiry 1985; 12:377-387. 12. Hanley JA, McNeil BJ: The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143:29-36. 13. Swets JA: Measuring the accuracy of diagnostic systems. Science 1988; 240:12851293. 14. Daley J, Jencks 5, Draper D, Lenhart G,

American Journal of Public Health 77

Iezzoni, et al.

Thomas N, Walker J: Predicting hospitalassociated mortality for Medicare patients: A method for patients with stroke, pneumonia, acute myocardial infarction, and congestive heart failure. JAMA 1988; 260:3617-3624. 15. Jencks S, Daley J, Draper D, Thomas N, Lenhart G, Walker J: Interpreting hospital

mortality data: The role of clinical risk adjustment. JAMA 1988; 260:3611-3616. 16. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: A severity of disease classification system. Crit Care Med 1985; 13:818-829. 17. Wagner DP, Knaus WA, Draper EA: Physiologic abnormalities and outcome from

acute disease: Evidence for a predictable relationship. Arch Intern Med 1986; 146:1389-1396. 18. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: An evaluation of outcome from intensive care in major medical centers. Ann Intern Med 1986; 104:410418.

.-.-.'.'.'.-.'.-.-.-.-.-.-..'.'...................:................................................:.:.:...-........... :::::" :.: .:.:...,.

'.'.'.'C.'.'.'.'.'.'.'.'.'.'.'.'... :.:.:.:.:.:.,. ................... .....e... ..................... .:.:.:.:-:.:.:.::.:.:.:.:.:.:-:. .............................. ...............-........... .................. ........:...:.:.:.:.:.:.:.:.:-, ::: :.:......e.....'e.. ...... :.:.:.:-, .:.:.:.:.:.:.:.:-:e :.:.,:.:.:-:.:.:.:-:-, ..... ............................ ...........................:........................... ... .. -:-:-:-:-:-::.,.-.............. ..........................................................C....e........................................... ..-:-: :.:.:.:-:.,.:.:.:.:.,- :.:. :.:....-:-. '.'.'.'.'.'.'.'.'.'.'.'.'.'.'........... ..........- :.:.:.:.:...:.:.:.:.:.:.:.....: ..................... .................................................. ............. .... :.:.... ............... .... ....... .......... .. ............. ............ ..................... ............. ............ .......... .. ...... :.:.:.:...'. ........... ..:.:.::.: .:.:.:.:.:.:.:-:.:.:.: :.:.:.:.:.:.,-.. ..-...-..........--.--...-..--..-..........-.-....-..............-..-................. ......... ..... :..:::.::::.:.:.,.,.::...:.:.:.::::.:::-.--'--.. ..:...:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.. :.:.:.:.:.... ....: :.:.:.... :.:.:.: :.: .:.:.....:-:.:.:: ::.: :.....:-,.,.,.,.,., '.. ........... ....:.:-:.: ........... ......... ..... :.:.:-:.:. .:...:.:-:.:.:.:.:.:.:.:-:...:.:-, i.:. :-, ..................... .............................. ............. ..................... ............. :......... ............. :...:.:-:.:.:.:.:.:.:........:.:.:.:.:.:.:.:.:.:.:.:-:.:.:.: :.:...:-:-, :-:-.-: :.:.:.-. ,:.:: :.:...:-, ................................ :: ::: :::....... :: ::.:.:.:.:.:.:.:...:-:-:.:. -'....... -:. .... ............ .......... ......... ................ ..................:.:.:.:.:.:.:.:.:.:.:.:.:. :. ..:.: .. .......... .:...:-:.:.:.:.:.:.:.:.:.:-:-:.:.......... .-.-.-.-.-:.:. :.:.: X. .:.:.:.:.:.:.:.:-, :::::::: :.:.:-:...:.:-:.:.:.:.:...:.:-:-, ;.:.:.:.:.:-:-:. :.:.:................-.-.-.-.'.'.-.'.'.'.-.. ................. ..... .:.:.:.: :.:.:.:-:.:.:.:.:.:.:.:.:.:. .......e.:.:...:-:.:.:.:.:.:.:.: :...:.:- .,.,.,::::.:::::::::::::::: ...................... :.:. ......:.::.:.:-.. ............ .................................:.:.:.:. .-:-:-.-:-:-:-:-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-. ....................... ......... ................ .:.:.:...-:-:-:-:-:-.-:-:.. .......-...-.-.-.-.-.:.:.:.:. .-:-:.:.:.:.:.:.:-:.:.:.:-:...: .... .................... ..............-.......-.......................... ..:..... .-.............. .%-.-.-.-.,-A%-..:.:...:.:.:.:.::. .... ... ........... .... ....................... .......:.:-::-:: :-.- ..-:.:.:.:.:.,.,..... ..... ...... :...:.:.:.:.:.::.:.: ............ ................... .................... ..... ............ ................ .... ................................................ :.:.:.:-:.:.,:.::.:.:....-:-,.,.,..,.,.., ................. ...................... .......... .... .......... ...............:.: ::...:-:.:.:.::.:.: :.:.... .................. :.:-:.:.:.:.:.:.:.:-:.,-.... ... .................. ..........................,....................... ....... .... ..........e.... .... ................................................ .............. ........... ............ ...... ...........,....... ............ .............. .:.:.:.,:.:.: :.:.:.:.:-:.:.:-:.:.:...:.:.:.:.::.:.: .... ............. .:.............. ........................................ :.:.: :.:.: :.:.:.:-:.:.:.:.....:-:.:.:.: :.:.:.:-:-...-:-,.. :-: ........... .... -..:,:.:.: .... :.-. .... ...........:...:.:.:..... :.:.:.:.:.:.:.:. .......... ..... ............... ........... ... ............... .... ................................... "-.-:.:.,..:-:.:.:.,e ........... ::.... -""" :.:.:.......... .............. .:.:.: .. .. ...................... P. :... ........... ......... .... ............i...................:.:.:.:.: :-;-:-X-:-:-:-:-: .........................-.--. ... %... ............ ..... :; ............ .. .;.a .Ir -W.- -.--.' .............. ............. .... .... .. .:..,.:.:.:.w-:.:.:.:.:.:.:.....:.:.:.. ...:.:....... :-::-: ...... .................... ...... .'e......

:.:.:.:-:.:.,-......

.::.::::::::: ...... ...........:............. . . . . . . . .;:::::::..................:: ::

:::::::::: .,.:.:.:.::::::::::::::::::::::,: :-:::

:::::

....:.:.:.:-,

::.:::::::-::-:

..:.:-:...:.:-,.-...:.:.:-:.:.::-:.:.:.:.:.:.:-:.:.:.::- :::::::::: -!:::.:.:.:::.:!ft-

......:.!..:.A............... .'..: ..:.. .:..-M..-.:.. ::::::::::

:..........-....

--------------------.-:-N:-:-X-:-:.:.

.:..

....

.....:.;.:.:....................................-.o

-:...:,................. :.::'.-.:.:.'.:.'.:.::::: .:.:.'.'.: .-. .!.--. W.1'.....:.:.:-:-, ......;....-:.:f: M.:.:........... ..... --------F-: .. ..i..---.--- .-!. .'.'-'.."IM.-Mm __-7

..:.:c..:-.-::-:-'-.:.:.:...I..

11.

1--l----.....

........ .... :I. ....%.::.:-:. ..............--''

.:%.:..

....-

:,:",

:.:.:-:-,:.:.:.:.:.:.:-:.:.:.:-:.:.

'::::::.:::::.:::::.::::.:.. ....

i

............:-:-:

...........

.............

..%,

..................

..% -.'.-..........:.:::............... ....... A. .-'...... ............. ......... ......-....

.......:....

.cM-:-:-:-:-:-:.:.:.

.............. :i.............. .......... .;.:.:.:.:-

'.

--.": ............-:-e .:.:.-.

.0.-.:. . . . . :.:.:.:.:.:.:.:.:.:. ....".11, iiii Ita"iffil, "i... *.F-."% :-..--"--'-'.--'..--'--

.... ..-W...bi

..:.

::!::::::::

..;.......... ::

.....'

......................

..................

.............

0-

................

..:.

:.....:. C!..... ................... .................. ...

...::::Umdmi!.--.---'! .''':.. ::

....

:.......

.....

...

.:.....

'.....Ie........I..........."...%...... .V,

.:M'f""-.-'o'_'-. K.!:]*.*]:.:.:

.....

:.:

.......I...... .: ....

.:MM..-'--'.-.'

............

:::...

....... .-.-.-.-.-.-.-. .......C............C.... ........

..... ........

............... -W............-...............

..-... ..:.:.:.,-,............. ......... .....:........ ..... -:

....... :.:.................. ...........'e.........

..

....:

.:.....

...............................

................ ..-. :-.......... ............ :........

..:.

I:........ ...... e.. ........ ..... :.---.....

............

..

:.:-:-"''

.e.-..;.:.:-, --ee... .....:::::::::::. ....-.............

..........................

FJ..... ..:.'.--: -461P m .....:.:. ,........ ... ..'--'W: .1:=:.lii.:..i. ....:.:.:.:.:.:. .:.:.:..

"'-

...................... ...

.:::::::-:::::::::::::;::::::::::::I::.::::;:;::: i"N...:.:.:....:.:.:.:-: :-:-. ..!.:...:.!..i!.v....

-....

:-

..-. ...

..'::::: ............

.....

..:-: :.:.:.:.:.:.:-:-, ..-X'-.-:-:-:-:-:-:

..................:.....:.:.:.

..................

-..:-:-:-:-.-.. ..W.:-. .::......

....... ............ ............. ....... ...... ..",...%.. ........

...:.:.:.:-:.:.:.:-..-.-...

.:.:.,.:.

....

...............

..............

.-.......

..".-.::.

..

..........

..

.:...:.:...... V...... ..ik.

..........:....... ...................

....

....

:.. .............' ........-......

..X... :...................

. . :.

..

__:

.................

..........:.:.::

.XI..:.:.:.:-:.:.:-:........ :.:. ..%............ %......

.............

... ..............

........... .-........... ................... ...:..................................... ...... .................. ............ ...:.:.:.:.:-:.:.:-:..........

I*.:-.-.-..-.-.........-.-.'.-.....::

:::

....... .-:............. .: .:-:.:.::.:. :-eup ..... .............. ... .. :.:.:.:-,. ..:.-.-.'.'.'.'.'.'.'.-.'..-.-.'.'.'.'.-.. ............... .................-... ............................................

P.W= MW:

..:.:.:...

-.4

...........................................

.:.................................................................................................::::

..

..

.:........ .......

................

...

...............................................................

... ......... .... ... ..:............................................................................................... :.. ................................ -:-:-:-::.:..,..... ...... ..:. :.:.:.:.:.:.:-:-:.:.::.:.:.:-:.:.,-:-:.,.::.:.:.:-:.:.:.:-:.:.:.:.:.:.:.:........ ......................

............................................................................ '..-.'.'.'.''.'.'.-.'.'.-.-.'...........

;:;.;:;ij..:;:;.:;:;: ..............::,..

... ,:---:.*.-.*:: ....2 ...:..................: ...............:...... ................. .-;.;-.-:-:-m,AV....I...................W, .'.'.'.'.-.'.'.-.-..'.'.'.'..... ........... .................. .:...................................... ..:. .:- .: :. :. ................. ,..-,..:..::.:.: ......:.

.-'-'..'

.:.:..

.:-:-,.. ........ ::.... .......... .... .........:...... -P.,.......:.:.:. ............ ........ .............. ....................-.......................... ;-:I....... ........ --... .....

...0"' ..'

...........::-::

.:.................

::.:.*.:.:.'-'::.lft

...:....... ..-.................-... e.......

:.:-:

:.:.:.

-.-..:.:.:.:.

................... :.:...:.:.:.:.:.:.:....... ..... ..--.- ....,W-: ..

... ...::..-.......::.:.:...,....

..,

. . . :.:.

.......... ............

.... "..:. .:c.:-, ..... .:.:.:.:.;.e .-: .---.-,A%--.-.-..................................

......... ........ .:.. -..... .::: .... ... ..... ................ .: .............. ................ ...:.-:.,-:.... ..:.:. ..........-.......... ..........

..:-:.

:.:.:.:.:.:.::.:.:.:-:.:. ....:.............:.:.:.:.:-:.:.:.:-, :.:.:-:-. ..:.:.:.:.:c. ..:................ :::::.::::,::::::::::-:: ............ ::::::::::: :.... .... ..........:. ................................

-................................................... :.:.:.:.:.:.:-:.:.:.:::.: .........

:.:.:-:-:.,-

.:.

.............%.................... ...'e ... -.'.-.-.-.-.-.'.'.'.'.'.'.'.-.-.'.......... ................................... .I................................................. .......................c... ................c.......................... .-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-..-.-.-.-................................... ...................... .................. c......e. ....-.................................................. .:... e..:... ...: ...........-....... .:.:.:.. !. .:.:.:.: . . . . . . . . . . . -. . .- .:-:. :: X..."............... --......... ................... -:::::::::.::.:.:-:.:.:.::.:.:-: ............,........-..............:.:.:.:.:.:.:.:.: .... :-:.:.,.,. ................ ............................. :-:-:-:- :-:-: --:::.:-:.:.,.:-:..:.::.:.:.:-:.:.............. :.:.:.:..' .. :: -:.:.:-:-,...... .'.'.'.'.'.'.'.'.................-...........-.. ...-.......................................................................... ........................................ ......-.......... '.'.'.'.'.'.'.'.'.'.'.-.'.'.'.'.'........... ... ..............:.:.:.:.:.:.: V................... .:.. ....: .%....................-.-.'.'.-.-............. .. ....... .............................................. ............................................ ...... '' .:-:.:.,......c..... .......!.:-:.:.,. .X -......... .... :.:.:.:.:.:-:-:.:.:-:.:.:.::.:...:....... ................... ......................... ..... ..... :...:-:.:.:::.:.-.... .......-...... ... 1..-. ...... .:..:.:A.- ..........:..... ................... ...... M.'..II ......:............ ......... ...........W............... -''-,-,-"b ................

::.-::

.....

....-.:-:.

..:.:.............................................................................

:"":::::

::;:::::::-, -...:....:.:.:.:.:.

..................;................: .::.... :-:-:.:.:..:.:.. :i:i:::.:.,.:,.-.:..i:...:i::.::::,-.,.::.::::: :::::::.:::::::.:.:.:.:--......

........W-ww. ..::.:.:.:-:-:.-....... -1",,--":,:,:,:,:,:",; ::::-::,:":, -,..:,..,'W...... ..-I..........'......... I;. W.-El":', ::: ...............".....Iil..............S.:.:.:.-.:.:.,:::::.:.:.::.:.:.:.:.

: : W.

""...............

..-........

.--W..--,.-."-,--", .".-.

........

.:.:............................................................................................................................ X.:...................

................ .... .......... ...............................................................................:.:...:.:.:.:.::.:.:.:.:...:-:.:.:.:.:.::.:.: ................... ...... ...... ...:-:.;.,.:-:.:.::.:.:.:.:-:.;.:-:.:.:.::.:.:.:-:....................... .: .: .: .......:...:.:-:.:. .:.:.:-:.:.:.::.: .::...:.:.:.:.:: .:.::.:.: .:.:..-.:.:.:.:.::.:.::.:.:.:.:.:.:.:.: ...................................:.:.:.:. .... -..-. ..... .....'..-...c............... ......... ........... ........... ... ............. ................. ............... ...................................- ................. ................... .... '...-.-.-.%-.-.':. .................................................................................... :.:.:.,.,.,-,.,.'' .................... ..............., .1---... .: .X .:.:.:.:-:-, ..... .'.'.'.'.'.'.'.'.....................................-..-.-.-.-............... . . . . -.-. . . . . . -.-. -. .-.- -............................... .:.:.:.:.:.. .:.:.:... :-:-,-:-:-, ..-. ::.::::..:.:.:.,.,::.:::.,.,-,.,.'' :-.......-................ ..:...:-:-,........................ .:... -.-C ... .....-:-:-:-:-:-::.:.::...:.:-:.:.:.:.:.:.::...:.:-:.:.:.:.:.:.:.:..... ...........'.'.'.'...................... :.:.:-:-:.,-.:...:-.-.......%-.....c...:.:. .............. .:-:.:.:.:.:.:.:-:.:.:.:.:.... ................................................................. ............ .. .....................:.:.:.:.:................................ :: .:-:.:.:.:.:.:.::.:.:.:-:.:.:.:.:.:.::.:.:.:-:.:.:-::.:.::.:.. :...:-:.:.:.:.:.:.:.::.,.:.:-: :. . ........ :. .:-:-, .::.:.:.:.:.:.:.::.:.:.:.:.:.:.:. ...... .......... ... :-:-:-:..:.:.:.:.::..:.:.:...:.:. .:-:.......... ......... ..................... .-.-.-.-.-..'.-.-.-.-.-.-.%......... .:. .. :.:.:.:.:.:.:-:.::.:::.:.:-:...--.:-:.:.:.:-:.:.:.::.,.:.:.:.:.:-:-.'-X-' ..... ......... :c ....................... .... -........................ ................... ........ ...... .. .......:. .:........... ................................................................... .......................................'.'.'."''..-...........-.......-..-.'......................................... ...'.-.-.-.%-.-.%-.-.... .... ......... ...................... .................. .................... ...-.-.-.-..-.. .......-.. .......................................................................................... ................................... .......,:::::!: :.. -:::-:-:-: ....................... ............................................................................................................................-......-................................................................................................................ ... ..'.-.-.'...-.-,.-.-.-..'.-.".,-.-.... ................ ...... --m-m-m....-............A. ...."-'.'.,-.'. :: ..::::::::::::::::::::::::: :::::i::::::::::: ...:... ,::::: :::: :::: :::::: :. ...::.... :..: ...:...:... ...:... X. ..::... ...

.--.-.--.%--.-.%-.-A.............. ....

.................................................................................................................................................. .:-:...,-:.:.::.:.,.:-:.:.:.:-:.:.:.::....:-....:.:-:.:.::....:-:.:.:.:.:.:.:.:.:..:-:-:.:.:-:.:.:.::.:.:-.-'.'.'.:.:.:.:.:.:.::.:.,.::...:.:.:.:.::.:.:.:-

::::::.::::.:::.::::.:.:::::::::::::::::::::.::.:::::::::

%...........:.:....

....:...... ...

........

:..-.........

78 American Journal of Public Health

........

.......X .::. :.:-:.:.:.:::N .:.:.:.:-:.:.,.:.:.:............. .... X.

:-'..

.-.-.-.%-.-....'.-.-.-.-.-.-.-.-..'.'....:........%-.-.'.'..:-.-.-.--;-X.:...

'-'-'-'-'-'.-.-.-.-.-.'.'.'...-.........

----I..

............................... ........:.:.:.:.:.:..:.:.:.:-:.:.:-:.:.:.:.:.:.:-: .......... ......................... ......

....::..:::.:.:.:.-.-'.::...........

------............. ................. ""'iz....... .......................... ..::::.. ............. ............................ ..-----------------------................ ..:.. :.:.,-.-,:.:-:.,.:.:.:.:....... .-...:.:.: ...... .......................................... .. .:.:.:.:.:-:-:.:.:-:...: ..... .::.:.::. .:....... ..:.........

I..0..

..:... .:.:.:.:.:.:.:.:.:.::e ..............:...............:: .:....-.%:.. .-..

--7-9--:-...................................

January 1991, Vol. 81, No. 1

Admission and mid-stay MedisGroups scores as predictors of death within 30 days of hospital admission.

We examined the ability of MedisGroups, a severity measure based on clinical data abstracted from the medical record, to predict mortality 30 days fol...
4MB Sizes 0 Downloads 0 Views