Identification of Potentially Avoidable Pediatric Hospital Use: Admitting Physician Judgment as a Complement to Utilization Review

(1)

Identification

of Potentially

Avoidable

Pediatric

Hospital

Use:

Admitting

Physician

Judgment

as

a Complement

to Utilization

Review

Jeffrey L. Soulen, MD”; Anne K. Duggan, ScIY; and Catherine D. DeAngelis, MD, MPH*

ABSTRACT. Objective. To determine, for acute gen-eral pediatric admissions, 1) the proportion of

admis-sions that the admitting physician would identify as

potentially avoidable; 2) services that might help reduce avoidable admissions; 3) potential savings in hospital days through eliminating avoidable admissions; 4) the

role that social factors play in admission decisions; and 5)

if admitting physician judgment provides a valuable complement to Utilization Review (UR) in identifying avoidable admissions.

Methods. A questionnaire was used to determine the

proportion of 600 acute general pediatric admissions con-sidered potentially avoidable by the admitting physician and to identify services required to prevent such

admis-sions. Savings through eliminating potentially avoidable

admissions and social factors influencing admission de-cisions were examined. Physicians’ and UR assessments

of necessity of inpatient care were compared.

Results. Admitting physicians judged 28% of

admis-sions as potentially avoidable, citing alternative services

for each avoidable admission. Eliminating all avoidable admissions would have reduced hospital days by 7.7%. Physicians identified 6% of admissions as social admis-sions, most commonly due to an overwhelmed family. UR identified 1% of potentially avoidable admissions as

unnecessary.

Conclusion. Incorporating admitting physician

judg-ment may significantly improve current efforts to

iden-tify and reduce avoidable hospital use. Pediatrics 1994;

94:421-424; hospital use, physician practice patterns,

physician decision making, utilization review.

ABBREVIATIONS. UR, Utilization Review; ER, emergency room;

JHCC, Thejohns Hopkins Children’s Center; CI, confidence

inter-val; SSU, short stay unit; LOS, length of stay.

The past decade has spawned considerable efforts

to identify and eliminate unnecessary hospital use

through Utilization Review (UR) and other

adminis-trative mechanisms.’3 Recent studies of the

effective-ness and cost of these policies suggest that such

efforts are no longer continuing to reduce hospital

use4 and have created significant new costs and

bu-reaucratic burdens.5 At the same time,

well-docu-From the of General Pediatrics, Dept of Pediatrics, Johns Hopkins

University School of Medicine, Baltimore, MD.

Presented in part at the annual meeting of the Ambulatory Pediatric Asso-ciation, New Orleans, LA, May 1, 1991. (Abstract published in AJDC.

1991;145:415-416.)

Received for publication Nov 9, 1992; accepted Feb 23, 1994.

Reprint requests to (J.LS.) The Sheppard and Enoch Pratt Hospital, 6501 N Charles St, P0 Box 6815, Baltimore, MD 21285-6815.

mented geographic variation in hospital use implies

that much hospital use is discretionary.6 These

ob-servations suggest that a broader approach to under-standing avoidable hospital use is needed if we are to identify its causes and provide effective, efficient solutions.

An important source of information on

discretion-ary

hospital use is one that is not used by current

mechanisms: the judgment of the admitting

physi-cian.

This

subject has been studied using hypotheti-cal case studies;7 to our knowledge, physicians’

judg-ment of their

own

“real life” admission decisions has

been evaluated only in one small study in 1976,

which found that 16% of adult admissions from an

emergency room (ER) were considered avoidable by the admitting physician.8

The purpose of this study was to determine, for

acute general pediatric admissions, 1) the proportion

of admissions that the admitting physician would

identify as potentially avoidable; 2) services that might help reduce avoidable admissions; 3) potential savings in hospital days through eliminating avoid-able admissions; 4) the role that social factors play in admission decisions; and 5) if admitting physician judgment provides a valuable complement to UR in identifying avoidable admissions.

METhODS

The unit of analysis was the individual admission. The study

was conducted at The Johns Hopkins Children’s Center (JHCC), a

large urban teaching hospital with inpatients both with common

pediatric problems and with medical and surgical problems

re-quiring subspecialty care. To focus on general pediatric patients,

admissions to surgical, psychiatric, and intensive care services

were excluded. Elective admissions were also excluded, as a large

proportion of elective admissions to theJHCC are for subspecialty

treatment protocols and therefore are not representative of a

gen-eral pediatric population. A sample size of 600 was selected to

provide a 95% confidence interval (CI) of ±2.5% for an estimate

that 10% of admissions would be considered avoidable.

The instrument was a dosed-ended questionnaire (available

from the authors upon request) developed from the ideas of 13

pediatric generalists and subspecialists and from the framework

provided by Mushlin and AppeL8 The questionnaire was refined through a 1-month pilot study of 113 admissions. The question-naire addressed three areas: 1) alternative services that, if available

and accessible, might have obviated admission; 2) social factors

influencing the admission decision; and 3) whether the admission

would have occurred in the absence of identified social factors. A

potentially avoidable admission was defined as one for which the

admitting physician identified one or more services that, if avail-able and accessible, might have eliminated the need for admission.

A social admission was defined as one for which the admitting

physician stated that the admission would not have occurred in

the absence of identified social factors. An admission influenced

by social factors was defined as one for which the admitting

(2)

admis-422 IDENTIFYING POTENTIALLY AVOIDABLE PEDIATRIC HOSPITAL USE

sion decision, but stated that the admission still would have

occurred in the absence of identified social factors.

All eligible admissions to the JHCC during the study period

were identified from daily admission lists. Based on the pilot

study, selection was limited to a two of three systematic random

sample of eligible admissions to include admissions over at least

three seasons. The study period was defined as consecutive days beginning in March 1990, exduding 2 weeks in July when the principal investigator OS.) was out of town, until October 1990

when data on 600 admissions had been obtained.

Because the majority of acute admissions to the JHCC are through the ER, questionnaires were placed in the ER. The ER

attendings and senior residents were asked to complete

question-naires when possible at the time of admission. Questionnaires not

completed at admission were delivered to the physician

respon-sible for the decision to admit and followed up until completed.

Data on UR were obtained from monthly hospital UR reports.

TheJHCC uses UR criteria similar to those used by many hospitals

and third-party payors (available from the authors upon request).

Admissions must qualify under an admission code (documenting

general reason for admission) and an acute code (meeting criteria

for inpatient care). Admissions are also assigned barrier codes

(documenting inpatient care that is inappropriate or unnecessary),

and social services related codes (documenting unnecessary

hos-pital days largely related to placement issues). In this study, an

admission judged unnecessary by UR was defined as one for

which the day of admission was judged inappropriate by UR.

RESULTS

There were 3032 admissions to the JHCC during

the study period. Of these, 2128 surgical, psychiatric,

intensive care, and elective admissions were

ex-cluded, leaving

904

acute medical admissions

eligi-ble for study. After randomization, 601 were

in-cluded in the study. Questionnaires were completed

for 600 of

601

admissions, with 41 % completed at the

time of admission, 72% within 24 hours, and 90%

within 3 days of admission. Questionnaires were

completed by 96 physicians. Fifty-five percent were

completed by residents, 45% were completed by

fac-ulty and fellows.

Study Population

Fifty-two percent of the study population was

male; 63% were black. Fifty-five percent of the

pa-tients were under 5 years of age; the rest were fairly

evenly distributed from age

5

to 21. Most (85%)

admissions were from the ER; 12% were from

hos-pital-based clinics, and 3% were from outside sites. Infectious ifinesses accounted for 41 % of

admis-sions. Leading infectious diagnoses were suspected

sepsis, ceilulitis, fever and neutropema, and

pneu-moma; each was responsible for

5%

to 6% of the

study sample. Respiratory ifinesses, primarily

asthma, accounted for 26% of admissions. Hemato-logic ifinesses, largely sickle cell disease, accounted

for 13% of admissions. Neurologic ifinesses

ac-counted for 5%, nutritional and metabolic disorders

accounted for 5%, and various other diagnoses

ac-counted for 10%.

Physician Responses - Potentially Avoidable

Admissions

Physicians identified 166 (28%) of the 600

admis-sions as potentially avoidable. The service most

fre-quently cited as an alternative to admission was a

24-hour short stay unit (SSU), for 23% of all admis-sions (Table 1). An outpatient department discharge

TABLE 1. Services Desired to Allo

tentially Avoidable Admissions

w Alternativ es to 166

Po-Service Admissions Percent of Percent of

Potentially 166 600

Avoided Avoidable Admissions

(No)* Admissions Studied

24-h short stay unit 138 83.1% 23.0%

Outpatient department 45 27.1% 7.5%

discharge coordinator to arrange available services

Home nursing services 29 17.5% 4.8%

Equipment or medication 13 7.8% 2.2%

to dispense

Sick day care 8 4.8% 1.3%

Transportation services 7 4.2% 1.2%

Chronic care placement 6 3.6% 1.0%

Foster care placement 5 3.0% 0.8%

Othert 9 5.4% 1.5%

Any service 166 100.0% 27.7%

* Many admissions had more than one alternative identified, so

the numbers in this column total >166.

t Three - magnetic resonance imaging at night; two - intensive

outpatient social work support; one each - weekend access to

subspedalty clinic, 24-hour home nursing, a different primary

pediatrician, and home aide to feed child.

coordinator who was able to help arrange services

known to be available was desired for 8% of admis-sions. Home nursing services were desired for 5% of admissions, and various services such as home

neb-ulization therapy, sick day care, transportation

ser-vices, and others were desired for another 8%. The fact that admitting physicians desired an outpatient department discharge coordinator to help arrange known services for 8% of the study sample suggests

that the

time

and effort

needed

to arrange services

presented a major barrier to using known services.

Potential Savings in Hospital Days

Admitting physicians desired services other than a

SSU as an alternative to 60 (10.0%, 95% CI 7.6% to

12.4%) of

600

admissions. These admissions

ac-counted for 212 hospital days, or 6.8% (95% CI 5.9%

to 7.6%) of the 3138 hospital days used by the study

population. There were an additional 30 admissions

for which a SSU was the only desired alternative and

length of stay (LOS) was only I day, yielding a total

of 90 (15.0%, 95% CI 12.1% to 17.9%) of

600

admis-sions that were potentially avoidable had all desired

services been available and accessible. These

admis-sions represented 242 (7.7%, 95% CI 6.8% to 8.7%) of 3138 inpatient days potentially saved. Estimating po-tential financial savings was beyond the scope of this study.

The 30 admissions for which a SSU was the only

desired alternative and

LOS

was only 1 day are

included in the above analysis as they likely would

have used no inpatient bed-days and would produce

a financial saving if a SSU day were less expensive

than an inpatient day. The 76 admissions for which a SSU alone was desired but

LOS

was >1 day were not

included as potential saving in hospital days from

these admissions is not possible to estimate.

at Viet Nam:AAP Sponsored on September 1, 2020

www.aappublications.org/news

(3)

Physician Responses - Social Admissions

Physicians identified 38 (6%) of 600 admissions as social admissions. An additional 8% were influenced by social context but

still

would have occurred

inde-pendent of social issues. The most common social

factor was an overwhelmed family, affecting about

half of social admissions (Table 2). Not trusting the

family also influenced about half of social admis-sions. Other social factors included not

knowing

the

family well enough to trust their judgment,

time

of day, transportation problems, and a few miscella-neous circumstances. Physicians were able to

iden-tify alternatives to admission for 30 (79%) of the 38

social admissions. For only eight of 38 social admis-sions did the admitting physician think that admis-sion was unavoidable.

Assessment by UR

UR identified

two

(0.3%) of 600 admissions as

medically unnecessary. Of the

166

admissions con-sidered potentially avoidable by the admitting phy-sician, UR judged two, or I

%,

as unnecessary. Of the

434

admissions considered unavoidable, UR judged none as unnecessary. Of the 38 social admissions, UR judged one as medically unnecessary.

DISCUSSION

Potentially Avoidable Admissions

This

study demonstrates that systematic attention to admitting physician judgment can provide

valu-able information about potentially avoidable

admis-sions that are not identified by UR. It is important to

understand that the issue is not simply to compare

the assessments of admitting physicians with those

of UR. In fact, the agreement was excellent between admitting physicians and UR in this study. The phy-sicians judged admission to be the most appropriate course of action, given available resources, in all 600 admissions, while UR judged 598 of 600 as appropri-ate. The

two

important questions are: 1)

does

poten-tially avoidable hospital

use

still

occur

even when

objective UR criteria are met? 2) might a broader

TABLE 2. Factors Influencing 38 Social Admissions

Social Factor Number of

Admissions*

Percent of 38 Percent of 600

Social Admissions Admissions

Family overwhelmed 20 53% 3%

Didn’t trust familyt 18 47% 3%

Didn’t know well 6 16% 1%

enough to trust

Timeofday 6 16% 1%

Distanceor 5 13% 1%

transportation

Courtesy to outside 3 8% 0.5%

physician

Othe4 2 5% 0.3%

Any social factor 38 100% 6.3%

* Many admissions were affected by multiple social factors, thus

the numbers here total >38.

t Reasons for mistrust included suspected abuse, parents known

to be noncompliant with medications, and one set of parents

known to give home intravenous therapy against medical advice.

: One decreased home nursing due to curtailed reimbursement,

one “great teaching case.”

perspective, inducing both objective criteria and

at-tention to admitting physician judgment, provide

data that would help to identify avoidable hospital

use and suggest means for reducing such use? The findings do not suggest that UR with objective

criteria should not be used. It is clear that UR has contributed to significant reductions in unnecessary

hospital care9 and that physician judgment is

influ-enced by so many complex and subjective factors that it cannot stand alone as a reliable, valid measure

of necessity.1#{176}

The findings do, however, suggest caution in

us-ing objective criteria alone to categorize hospital use

as medically necessary or medically unnecessary. In

this study, physicians frequently

judged

outpatient treatment medically feasible, yet chose inpatient treatment for complex reasons involving the context

in which a particular illness occurred. UR was not

sensitive in identifying the scope of such

discretion-ary admissions. We would do well to learn more

about what might allow discretionary admissions to be replaced by outpatient alternatives. Admitting

physicians identified deficits in services which, if

addressed, could result in decreased hospital use.

UR, by its design,

does

not provide such information.

In this population, admitting physicians identified

28% of admissions as potentially avoidable, a

consid-erably larger proportion than the 16% of adult

ad-missions found by Mushlin and Appel.8 Also

note-worthy was the physicians’ eagerness to provide this information, demonstrated by the 99.8% response

rate. The information provided by physicians about

services that would help to reduce avoidable admis-sions and about barriers to using such services was specific and quantitative. Particularly helpful for this

population would have

been

a

24-hour SSU, help

arranging known services, and prompt access to

home nursing.

Social Admissions

This study also demonstrates the considerable role that social context plays in admission decisions. Phy-sicians identified 6% of admissions as social

admis-sions, listing the factors leading them to decide on

admission and delineating services that would have helped prevent the majority of these admissions. UR

was not sensitive in identifying potentially avoidable social admissions. The medical system is still relied

on to address social needs; the results reflect the

failure of the social system to provide these needs, not the failure of physicians caring for families or of UR in reviewing those admissions.

Study Limitations

This study has a

number

of limitations. Most

im-portant, we were unable to address whether

physi-cians actually would have used the services they

cited as alternatives to admission. A reduction in

hospital days cannot conclusively be demonstrated

short of making such services available and accessi-ble.

Second, the population at an urban teaching

hos-pital is not representative of the population as a

(4)

424 IDENTIFYING POTENTIALLY AVOIDABLE PEDIATRIC HOSPITAL USE only nonelective general pediatric admissions, which

resulted in a population more typical of national

pediatric hospital use. It would be valuable to collect similar data for pediatric elective and surgical admis-sions, for adults, and in a variety of geographic set-tings.

Third, this study does not address whether finan-cial cost would have been reduced had these patients

been

treated as outpatients. Several studies have doc-umented decreased costs with a shift from inpatient to outpatient care.9”2 One recent study, in contrast,

has projected that future savings in inpatient costs

may largely be offset by increases in outpatient

costs.4 There are, however, advantages to outpatient care that cannot be measured simply by comparing

charges for services. Every admission that is avoided

reduces potential morbidity, mortality, and cost

cre-ated by nosocomial infections and iatrogenic

compli-cations.13 Recovering at home in a familiar environ-ment is a benefit. Reducing delays in care due to lack

of available inpatient beds is an important goal for

crowded facilities.

Policy Implications

The

findings

of

this

study have policy implications

in a number of areas. The most immediate is the need

to identify the true scope of avoidable admissions, to identify services that could help prevent avoidable

admissions, and to make such services available and

accessible. More fundamentally, we must consider

the possibffity that our current paradigm for identi-fying avoidable hospital use, which relies on objec-tive criteria, may represent an incomplete approach. With alternatives that are unique to each community

and constantly changing,

with

a

myriad of social

factors and other intangibles that are unique to each

patient, the decision to hospitalize is ultimately a

subjective one. Although further refinement in using

objective criteria is essential, it may not be wise to be

too single-minded in an effort to reduce hospital

admissions with objective criteria for medical neces-sity. A paradigm that includes both objective criteria

and admitting physician judgment may be much

more useful in identifying the causes of avoidable

admissions and in identifying effective alternatives. Finally, reducing avoidable hospital use involves changing physicians’ practice patterns. Research in modifying physicians’ practice patterns has identi-fled

six

general approaches: education, feedback,

ad-ministrative rules, financial incentives, penalties, and direct participation by physicians in the efforts to

change practice patterns.’0 Combinations of

ap-proaches work

better

that any single approach.

The current system applies rules, incentives, and

penalties to hospitals, which in turn apply education and feedback to physicians. Limited reimbursement for outpatient services continues to provide financial incentives to hospitalize. Participation by physicians,

which has been shown to be a key coi.nponent in

some previous attempts to modify hospital use,1#{176}is conspicuously absent.

CONCLUSION

Systematic attention to admitting physician

judg-ment is a valuable adjunct to UR in identifying

avoidable hospital use while providing information

about services needed to reduce such use. Progress

in reducing avoidable hospital use may result from increasing physician participation.

REFERENCES

1. Gertman PM, Restuccia JD. The appropriateness evaluation protocol: a

technique for assessing unnecessary days of hospital care. Med Care. 1981;19:855-871

2. Siu AL, Sonnenberg FA, Manning WG, et aL Inappropriate use of hospitals in a randomized trial of health insurance plans. N EnglJ Med.

1986315:1259-1266

3. Kemper KJ. Medically inappmpriate hospital use in a pediatric

popu-lation. N Engl Med. 1988318:1033-1037

4. Schwartz WB, Mendelson DN. Hospital cost containment in the 1980s:

hard lessors learned and prospects for the 1990s. N Engi IMed. 1991; 324:1037-1042

5. Woolhandler 5, Hinunelstein DU. The deteriorating administrative

ef-ficiency ofthe US. health care system. N Engi JMed. 1991324:1253-1258

6. Perrin JM, Homer CJ, Berwick DM, Wooff AD, Freeman JL, Wennberg JE. Variations in rates of hospitalization of children in three urban communities. N Engl IMed. 1989320:1183-1187

7. Kuder JM, Vilinain JA, Demlo LK. Exploring physidan responses to patients’ extramedical characteristics: the decision to hospitalize. Med

Care. 198725:882-893

8. Mushlin A!, Appal FA. Extramedical factors in the decision to hospi-talize medical patients. Am JPublic Health. 197666:170-lfl

9. Wickizer TM, WheelerJRC, Feldstein PJ. Does utilization review reduce

unnecessary hospital careand contain costs? Med Care. 198927:632-647

10. EisenbergJM. Doctors’ DeciSiOns and the Cost ofMedical Care. Ann Arbor,

Ml: Health Administration Press; 1986:-86,99-142

11. Feldstein PJ, Wickizer TM, Wheeler JRC. Private cost containment: the

effects of utilization review programs on health care use and

expendi-hires. N Engl IMed. 1988318:1310-1314

12. Wickizer TM, Wheeler JRC, Feldstein PJ. Have hospital inpatient cost containment programs contributed to the growth in outpatient

expen-ditures? Med Care. 199129:442-451

13. DeAngelis C, Joffe A, Wilson M, Willis E. latrogenic risks and financial costs of hospitalizing febrile infants. AJDC. 1983;137:1146-1149

(5)

1994;94;421

Pediatrics

Jeffrey L. Soulen, Anne K. Duggan and Catherine D. DeAngelis

Judgment as a Complement to Utilization Review

Identification of Potentially Avoidable Pediatric Hospital Use: Admitting Physician

Services

Updated Information &

http://pediatrics.aappublications.org/content/94/4/421

including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or in its

Reprints

(6)

1994;94;421

Pediatrics

Jeffrey L. Soulen, Anne K. Duggan and Catherine D. DeAngelis

Judgment as a Complement to Utilization Review

Identification of Potentially Avoidable Pediatric Hospital Use: Admitting Physician

the World Wide Web at:

The online version of this article, along with updated information and services, is located on

been published continuously since 1948. Pediatrics is owned, published, and trademarked by the

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has

http://www.aappublications.org/site/misc/Permissions.xhtml

http://www.aappublications.org/site/misc/reprints.xhtml