Neonatal
Hotline
Telephone
Network
Paul H. Perlstein, MD, Neil K. Edwards, MS PE, and
James M. Sutherland, MD
From the University of Cincinnati Medical Center, Crosley Memorial Nursery, Cincinnati General Hospital, Cincinnati Children ‘s Hospital Medical Center and the Children’s Hospital Research Foundation, Cincinnati
ABSTRACT. By simplifying the process by which
tele-phone contacts are made, improved communications were
established between a university-affiliated newborn in-tensive care center and some of the community hospital nurseries that it serves as a regional resource. Initiation
of the improved system of communications was
associ-ated with a significant improvement in the survival of
infants transferred from the community hospitals to the
regional care facility. Pediatrics 64:419-424, 1979; corn-munications, regionalization, rnortality.
Since 1968, the Cincinnati Children’s Hospital
Newborn Special Care Unit (NBSCU) has been a
southwest Ohio regional facility to which more than
7,000 sick newborn infants have been transported
for intensive observation and care. Since their
in-auguration, the newborn intensive care facilities in
Cincinnati have always been very well used by the
community; but soon after the NBSCU was
estab-lished it became clear that, except for the
conver-sations needed to physically implement an infant
transfer, dialogues between personnel in the
refer-ring and referral centers were almost nonexistent.
Beginning in 1972, with assistance from the Ohio
Department of Health, an attempt was made to
bridge a possible gap in communication by initiating
a program to install direct, unswitched hotline
tel-ephones connecting some of the community hospi-.
tat nurseries to the Children’s Hospital NBSCU. It
was hoped that these phones would enhance the
sense of community within the regional network of
hospitals and make the consultation services in the
NBSCU more available to those responsible for
Received for publication Dec 27, 1978; accepted Feb 20, 1979. Reprint requests to (P.H.P.) University of Cincinnati, College of Medicine, 231 Bethesda Ave. Room 6103, Cincinnati, OH 45267.
PEDIATRICS (ISSN 0031 4005). Copyright © 1979 by the American Academy of Pediatrics.
evaluating and delivering care to high risk infants not yet transferred to the special care regional nursery.
In the following report, this hotline telephone
system is described along with the measures used
to evaluate its effect as a device for bringing regional
medical resources closer together.
METHODS
Between 1972 and 1975, telephones were installed
in six different hospital nurseries and connected by
direct lines to a special phone located in the
NBSCU at the Cincinnati Children’s Hospital.
Each of the six hospitals was located in a
five-county southwest Ohio region within a 50-mile
ra-dius of Cincinnati. Since the direct wires precluded
the use of switching equipment, no dialing was needed and the personnel in these community
nur-series had only to lift the phone off its cradle to
establish an immediate contact with the Cincinnati
special care nursery. At the Cincinnati Children’s Hospital, the phone was answered by clerks, nurses, pediatric housestaff, neonatology fellows, or staff
neonatologists. If the person answering the phone
did not feel qualified to deal with the particular call
received, the call was referred to an appropriate
resource person.
The NBSCU nursery personnel were encouraged to use the telephones to exchange information
about procedures and principles important to
pro-viding optimal care to newborn infants born in the
participating hospitals. The phones also were used
to help facilitate the transfer of sick infants from
their hospitals of birth to the Children’s Hospital
Regional Intensive Care Unit, and to keep
person-nel and parents remaining at the hospital of birth
informed of the clinical condition and progress of their babies after transfer. To maintain records of
information exchanged, an approved recording
ma-chine with an audible tone generator was attached
to the phone system and all telephone calls were electronically monitored.
Changes in neonatal mortality were used as the
major measure in evaluating the system’s impact
on newborn care. Three of the six hospitals served
by the hotlines were selected as the sources of study
data. These three phone-connected study hospitals
were selected from among the six simply because
they were not university-affiliated. In addition,
dur-ing the years of study, these hospitals were
gener-ally immune from major changes in policy,
person-nel or other influences, such as expansion in the use
of fetal monitoring, which would have made it
dif-ficult to separate an effect of the telephone from
those resulting from coincident forces. Conversely,
the three phone-connected hospitals not included
in this analysis were eliminated because of their
university affiliations and identifiable multiple
changes in infant care practices associated with and
including the fact that each had a neonatologist in
charge of their nursery during the study period.
Four other non-university-affiliated hospitals
which did not have access to the hotline phone
service during the study years are located within 50
miles of Cincinnati. Of the four
non-phone-con-nected hospitals, three were used for tabulating
control data with which the study data could be
compared. These three hospitals had obstetric and
newborn services of the same size as the three study
phone-connected hospitals and similarly use the
Children’s Hospital as a regional referral center for their high risk infants. The eligible hospital not
used as a source of control data was eliminated
because of its small obstetric service which delivers
less than 300 infants each year.
In summary, the hospitals eliminated from this
analysis were eliminated strictly because of
over-whehning influences that interfered with making a
clear separation between the effect of the telephone
installation and the effects of other variables.
The raw data upon which the analyses were
performed were obtained from the Ohio
Depart-ment of Health, the Cincinnati Bureau of Vital
Statistics, the American Hospital Association, and
from each of the studied hospitals.
SUBJECTS
There was a total of 45,000 babies born in the
three study and three control hospitals in the years
1970 through 1974. The individual hospital births
ranged between 1,100 and 2,500 births in 1970 with
a gradual decline in births recorded in each of these
hospitals between 1970 and 1974. This decline was
consistent with and representative of the changing
birth rates in all of Ohio during these same years.
The percentage of infants born weighing less than
2,500 gm in these six hospitals ranged from 6% to
8% and in none of the hospitals did the percentage
of low birth weight infants change with any
statis-tical significance in the years 1970 through 1974.
Additionally, there was no statistically significant
yearly variation in the 6% to 8% of nonwhite infants
delivered in any of the study or control hospitals
during this period of time.
In expressing the results, no correction has been
made to account for interhospital variations in the
hospital specific baseline percentages of low birth
weight or non-white infants delivered. When made,
these corrections do not alter the significance of the
results. Because of difficulty in obtaining the
nec-essary information, the mortality figures are not
limited to infants who died within the first 28 days
oflife but include all liveborn infants who died prior
to discharge from the hospital care occasioned by
their births. In all of the hospital specific tabula-tions, the deaths of infants after their transfer to
the NBSCU were reassigned to their hospitals of
birth. Absolute mortality rates are expressed as
deaths per 1,000 live births.
RESULTS
The yearly neonatal mortality rate for infants
born in the first study hospital varied insignificantly
between 12.0 in 1970 to 13.1 deaths per 1,000 live
births in 1972. At the end of 1972, a hotline
tele-phone was installed in the hospital’s nursery.
In 1973 (the year following the installation of the
telephone), the mortality rate ofinfants born in the
first study hospital, as tested by analysis, dropped
a significant 21% from 13.3 to 10.4 (P
<
.01). Asifiustrated in Figure 1, this 21% fall in mortality rate
also was significantly different (P
<
.01) from the15.4 to 14.4 fall in mortality rate that was registered
by infants born in the first non-phone-connected
control hospital, the 127 to 12.1 change recorded
for infants born in the second control hospital, and
the stable mortality rate of 9.8 tabulated for infants
born in the third control hospital during 1972 and
1973.
Between 1970 and 1974, the mortality rate
changes that occurred in the second and third study
hospitals were relatively stable and similar to those
recorded in the three control hospitals and shared
with the control hospitals the same significant
dis-similarity when compared to the first study hospital
change that occurred in 1973.
In 1974, the mortality rate for infants born in the
first study hospital declined another 12% to 9.1 per
1,000 live births. This continuing decline in
cz\
\ 6
5
4
3
2
II
0
9
8 7
6
- study hospitals -- control hospitals
MEAN DEAThS
PER 1000 LIVE
BIRTHS
YEAR
BEFORE YrAR YEAR YEAR
PHONE AFTER BEFORE AFTER I Lived 4982 4906
F
Lived 3474 3451Jr
Died 55 40 pied 42 4Lj[
Toto No 5037 4946 Total No 3516 3492p(.0O5
I
N.S. ] Fig 2. Mean mortality rates of babies born in study andcontrol hospitals before and after phones installed in
study hospitals.
.25
-I SE
.20
-
____
MEAN CHANGE IN % OF
LIVEBORN INFANTS I 5
-TRANSFERRED BEFORE & AFTER
PHONES AVAILABLE .1
.05
-PHONE NO PHONE STUDY CONTROL
HOSPITALSHOSPITALS
DIFFERENcE NOT SIGNIFICANT Fig 3. Change in infant transfer rates.
An attempt was made to elucidate the
mecha-nisms by which the presence of direct line
tele-phones enhanced survival. Prior to the installation of phones, a mean of 2.7 ± 0.6% SE of babies born
in the three study hospitals were transferred for
DEATHS care to the NBSCU. Following the phone
installa-PER 000 tions, this proportion remained stable at a mean of
LIVE BIRTHS
2.4
± OM%SE. A
similar constancy was found inthe percentage of babies transferred out ofthe three
control hospitals during the same time periods. The
phone installations, therefore, did not coincide with
any significant change in the number of patients
transferred out of either the study or control
hos-pitals (Fig 3). In addition, after the phones were
I I I I I
970 1971 972 973 974
s=phone installed
Fig I . Yearly mortality rates of infants born in each
study and control hospital.
mortality rate tabulated for infants born in the
three non-phone-connected control hospitals. In
1974 (the year after hotline phones were installed
in both the second and third study hospitals’
nur-series), the mortality rate of infants born in the
second study hospital was 7.5 which was a
signifi-cant (P
<
.01) decrease of 28% from the 10.4 deathsper 1,000 live births tabulated for infants born in
that same hospital in 1973. Likewise, the 6.1
mor-tality rate for infants born in the third study
hos-pital in 1974 was a significant 36% less than the 9.5
___________________________
per 1,000 livebirth rate tabulated for infants born in
1973.
__________
__________
Combining the mortality data for the three study
hospitals, an annual mean rate of 10.9
±
1.1 SE per1,000 live births was tabulated with data generated
by infants born before telephones were installed in
each of the hospitals. This pre-phone installation
death rate was not significantly different from the
12.3
±
1.9 SE death rate tabulated for infants bornin the same years in the three control hospitals.
This contrasted significantly, using a paired t test
(P
<
.005), when compared to the annual meandeath rate of 7.6 ± 0.9 SE per 1,000 live births computed with data on infants born in each of the
three study hospitals after the phones were
in-stalled. In contrast, a comparison of the combined
annual mortality data on infants born during the
same years in the three non-phone-connected
con-trol hospitals demonstrated no significant differ-ence between the mean of 123
±
1.9 SE and the12.5 ± 1.1 SE per 1,000 live births, respectively,
generated before and after phones were installed in
the study hospitals. These combined data are
10
F Iter
PHONE NO PHONE
STUDY CONTROL HOSPITALS HOSPITALS
made available, there was no decrease in the rate of
neonatal deaths that occurred in the hospitals of
birth before infants were transported to the
NBSCU.
Whether the transfer of infants occurred before
or after the phones were installed, however, did
make a difference in the ultimate survival of those
infants who were admitted to the NBSCU. Of
ba-bies transferred from the study hospitals in the year
prior to the phone installation, 17.3
±
3.6% SE diedbefore discharge from the NBSCU. With paired t
test, this was a significantly higher death rate than
the 12.9
±
3.1% SE rate tabulated for infantstrans-ferred from those same three hospitals in the year
following the initial access to the phone service.
This contrasted with the statistically insignificant
difference in the mortality rates of infants cared for
in the NBSCU after transfer from the three
non-phone-connected control hospitals during the same
time periods (Fig 4).
The enhancement in the survival of infants from
the study hospitals was coincident with an
associ-ated increase in the percentage of infants who were
transferred because of clinical decisions arrived at
more quickly following the installation of the
hot-lines.
In the following analysis, a two-hour maximum
was used as a dividing time in tabulating the ages
of infants at the time the decision was made to
transfer each to the NBSCU from each of the study
and control hospitals in the years 1972, 1973, and
1974. This two-hour dividing time was selected after
it first was determined that in 1974 all infants with
respiratory distress syndrome born in local
Cincin-% OF
TRANSFERRED INFANTS
THAT DIED
Lp<02J
Fig 4. Mortality rates of infants after transfer from
hospitals of birth to Newborn Special Care Unit
(NBSCU). Comparison of rates before and after phones
available in study hospitals.
nati hospitals and cared for in the NBSCU who
survived were transferred to the NBSCU at a mean
age of 2.7 hours. This contrasted significantly when
compared to the 9.7-hour mean transfer age of
Cincinnati-born infants with respiratory distress
syndrome who died after arrival at the NBSCU.
Since approximately 45 minutes is the minimum
time needed to transfer a baby from a local
Cincin-nati hospital to the NBSCU, it was presumed that
the decision to transfer the surviving infants was
made by 2 hours of age.
In 1972 (the year prior to the installation of a
direct line telephone), 23 infants were transferred
from the first study hospital. Only in one, or 4.3%,
of the 23 transfers completed, was the decision to
send the infant to the NBSCU made by the time
the baby was 2 hours of age. In 1973 (the year after
the phone was installed), the two-hour decision
time was met in the cases of six, or 33%, of the 18
infants transferred from the first study hospital.
This increase in the number of infants transferred
because of more rapid decisions was significant at
P < .05. In 1974 (the second year after the phone
was installed), the percentage of infants transferred
because of early decisions remained stable at the
significantly higher level of 35% of all transfers
made.
Of the 76 decisions to transfer infants from the
second study hospital to the NBSCU in 1972, 18, or
24%, were made by the time the infants were 2
hours old. In 1973, a similar 12, or 22%, of the 55
infants had their transfers initiated by 2 hours of
age. Both 1972 and 1973 were years prior to the
installation of a hotline telephone in the second
study hospital. In 1974 (the year after a phone was
installed in the second study hospital), a
signifi-cantly greater (P < .05) 40% of the 58 infants
transferred from the second study hospital were
transferred in response to decisions made by 2 hours
of age.
There were 46 infants born in the third study
hospital and transferred to the NBSCU in 1972. A
decision was made by two hours to transfer seven,
or 15%, of these 46 babies. In 1973, the proportion
of babies transported because of rapid decisions
remained stable at 14%, or eight out of 59 infants
transferred. At the end of 1973, a direct phone was
installed in the third study hospital and,
subse-quently, in 1974 when 44 babies were transferred to
the NBSCU, a significantly greater (P < .05) 32%,
or 14 infants, were transferred because of decisions
made by the time the patients were 2 hours old.
Therefore, paralleling the improvement in survival
that followed phone installations in each of the
study hospitals, a significant increase was tabulated
37.0
.) 36.
35.0
% OF
TRANSFER DECISIONS.
MADE BY 2 Hrs
AFTER BIRTH
P972 913974 1972 973 t974 972913974
r-i_] r1 Efl rIL
NS IIIIik_] C.05 (05 CONTROL FSTDJ ISTUDY L_!p!m Fig 5. Change in transfer decision time before and after phones available in study hospitals.
the phone service became available than was
tab-ulated for transfers made in prior years.
In contrast, annual transfers of infants from the
three control hospitals during this same period of
time remained essentially constant in the
propor-tion of transfers initiated by the 2-hours-of-age
breakpoint time. In 1972 this proportion was 22%;
in 1973, 22%; and in 1974, 19%. These data are
summarized in Figure 5.
Finally, coincident with the significant increase
in the speed with which decisions were made to
transfer infants after the phones were installed
when compared to the pre-phone years, there was
a significant increase (P < .02) in the mean admis-sion rectal temperatures ofstudy infants transferred
after the direct phone service became available (Fig
6).
DISCUSSION
That good communication plays a role in any
effective system of regionalized medical care is
probably so evident that acceptance of its role does
not depend necessarily on proof of its importance.
There are so many examples, however, of errors
being promulgated by solely trusting in the inherent
and obvious virtue of a technique, that it is
reassur-ing to find statistical support for a practice that
might have had no effect beyond simply being
time-consuming and expensive.
At least one other hotline system has been
de-scribed as a resource for improving care in a regional
perinatal system.’ The system described in this
paper, therefore, is not unique. This present study,
however, provides some insight into how telephone
communications can be enhanced to the point of
demonstrably improving the survival of high risk
infants whose management requires transfer to a
specialized center for newborn intensive care.
Since regular dial phone service was available
before the direct hotline phones were installed in
Before After
PHONE
Fig 6. Mean rectal temperatures ofinfants born in study
nurseries on admission to Newborn Special Care Unit
(NBSCU).
the study hospitals, the fact that the hotline
tele-phones were connected by direct wires and required
no dialing to call the NBSCU leads easily to a
conclusion that telephone communications can be
improved by simplifying the process by which
tel-ephone calls are initiated. In addition, the very
presence and clear single purpose of the hotline
telephones in the community hospital nurseries
must be presumed to have helped overcome at least
some of the inertia in the processes by which
trans-fer decisions are reached. That this decision process
is important in reducing neonatal mortality is
sug-gested by the observation that the enhanced
sur-vival associated with the activation of the hotline
telephone service coincided with an acceleration in
the transfer decision and subsequent transport of
infants from the phone-connected community
hos-pitals to the regional NBSCU.
It is also realistic to presume that the complete
cause and effect mechanisms are far more complex
than suggested by the simple statistical
relation-ships demonstrated. For example, review of the
tape-recorded telephone conversations that were
monitored during the study period produced
nu-merous examples of the telephone being used
infor-mally and effectively to teach community hospital
personnel how to quantify their assessments of
ba-bies’ conditions. Undoubtedly these informal
les-sons were aids in enhancing the speed with which
infant transfers were initiated.
During the years of this study, an average of 21
calls a day were handled by the hotline telephone
system. This approximates 13 telephone calls per
year per patient served. The average call was about
served, approximately 32 minutes of total personnel
time was expended during these conversations.
Both personnel time and hardware expenses were
used in making a cost analysis of this study which
leads to a final observation. if bified directly to
patient care, this telephone service would have
in-creased the cost of each patient’s hospitalization by
less than 0.13%. This seems a small price to pay for
the demonstrated coincident 30% improvement in
infant survival and the unquantifiable but immense
improvement in the quality of relationships that
now exist between the regional Cincinnati special
Infant care center and the greater community that
it serves.
ACKNOWLEDGMENTS
This work was supported in part by Department of
Health, Education and Welfare Grant MCR 390290 and
Ohio Department of Health Grant 731.
The authors acknowledge the cooperation, support,
and advice of all physicians, nurses, and parents who
have participated in making the system described for
communication work to the benefit of the babies born in
the region served.
REFERENCE
1. Baum RS: Consultation by newborn hotline. J md State Med Assoc 66:317, 1973.
THE PRODUCTIVE IDLENESS
Why do intelligent and competent Americans who have spent many years in
a series of reputable educational insitutions often appear to English people to
be fundamentally uneducated?
I think it’s because they work too hard.
From an early age they have spent so much time writing term papers, studying
for tests and pursuing independent research projects, that they have never had
a moment to themselves.
As a result (I must say this even if it sounds insufferable), they lack that sense
of measure, that instinct for what is of value in the life of the mind, that only
comes with years of productive idleness. I mean by “productive idleness” time
spent in undirected reading, uninterrupted thinking, and aimless talk. It is in
these idle hours that an intelligent person becomes an educated person.
Submitted by Student