Effect of Remote, retroactive intercessory prayer on outcome in trauma patients
Dear Sir,
We did a statistical appraisal of whether remote retroactive
intercessory
prayer affects in-hospital mortality or hospital length of stay after
hospital
admission for trauma. The intercessory prayer was done 1 month to 8 years
after the patients’ hospitalisation. The hypothesis was that remote,
retroactive intercessory prayer shortens hospital stay and lessens
mortality
after trauma.
Our medical centre’s institutional review board approved this study.
All patients recorded in the trauma registry of a Level 1 trauma
centre
between January 1996 and May 2004 were included in this study. The trauma
centre serves a metropolitan area of greater than 3 million inhabitants
and
has approximately 4,000 trauma admissions per year.
In June 2004 the patients were stratified into subgroups using
bifurcation
values for each of four variables known to have a strong impact on
mortality
after trauma. The variables were age, systolic blood pressure on arrival,
Glasgow Coma Score on arrival, and Injury Severity Score. Their lower
grouping values were age less than or equal to 55 years, systolic blood
pressure on arrival less than or equal to 100mm Hg, Injury Severity Score
less
than or equal to 8, and Glasgow Coma Score on arrival less than or equal
to
8. Within each subgroup, a block randomization into two groups was
performed using a pseudo-random number generator.
Once randomisation was complete the two groups of patients’ names
were
printed out, using the first name and last initial. A coin toss was used
to
designate the study group. The list of names for the study group was given
to
an intercessor who prayed for the health and well being of the people in
the
study group.
Two outcome variables were measured: length of hospital stay, and in-
hospital mortality.
The Mann-Whitney test was used to compare the study and control
groups to
verify they were comparable with regard to age, Glasgow Coma Score on
arrival, systolic blood pressure on arrival, and Injury Severity Score.
For these
variables, a 2-sided analysis was done based on a null hypothesis of no
difference between the two groups.
A 1-sided Mann-Whitney test was used to compare the study and control
groups’ hospital lengths of stay. The test was based on a null hypothesis
that
the stays of members of the control group tended to be shorter.
Fisher’s exact test was used to compare the study and control group’s
mortality rates. This test was 1-sided, where the null hypothesis is that
the
mortality rate of the control group does not exceed that of the prayer
group.
Our sample size of over 33,000 patients was sufficient to detect a 1%
difference in mortality rate, if such a difference existed, with 99% power
with
a Type I error of .05.
Of the 33473 trauma registry patients, 16736 were randomized to the
control
group and 16737 were randomized to the study group. There were no
significant differences between the groups with regard to age, systolic
blood
pressure on arrival, Glasgow Coma Score on arrival, and Injury Severity
Score.
Median length of stay was 2 days for the control group (range 0 to
375 days)
and 2 days for the study group (range 0 to 730 days). The p-value for the
associated test of hypothesis was .27. In-hospital mortality for the
control
group was 4.6% and 4.6% for the study group (p=.44).
Our study’s findings conflict with those reported by Leibovici (1)
who showed
that remote retroactive intercessory prayer was beneficial. It may be
argued
that the outcomes we studied – mortality and hospital stay – were not
appropriate measures to assess improvements in “well being”. It is
arguable
that, in some cases, death may end suffering, and so bring a form of
benefit
to a patient. Similarly, a shorter hospital stay may not necessarily
reflect
improved health or well-being. Additionally, we did not assess the long-
term
health outcome or the patient’s sense of well-being. It is possible there
were
differences in the two groups that we did not assess. Nor did our study
make
an effort to control for prayer that may have been said on the patients’
behalf
by other people, such as family members or friends. However, our aim in
this
study was to assess objective and measurable outcomes that are important
to
clinicians. Our thought (and perhaps it is a biased one) was that survival
was
better than death, and that a shorter hospital stay was better than a long
one.
Although we cannot claim we have shown remote retroactive intercessory
prayer provides no benefit to trauma patients, we think we can claim to
have
shown it is quite unlikely that it affects in-hospital mortality or length
of stay.
This study does not prove that intercessory prayer has no benefits. There
are
several compelling studies which appear to show the benefits of prayer,
and
we feel further research in this area is warranted.
1. Leibovici L. Effects of remote, retroactive intercessory prayer on
outcomes
in patients with bloodstream infection: randomised controlled trial. BMJ
2001,
323:1450-1
Competing interests:
None declared
Competing interests:
No competing interests
23 June 2004
Adam Starr
Associate Professor of Orthopaedic Surgery
William Frawley, Michael Bolesta, Christopher Espinoza-Ervin
University of Texas Southwestern Medical Center, Dallas, Texas, USA
Rapid Response:
Effect of Remote, retroactive intercessory prayer on outcome in trauma patients
Dear Sir,
We did a statistical appraisal of whether remote retroactive
intercessory
prayer affects in-hospital mortality or hospital length of stay after
hospital
admission for trauma. The intercessory prayer was done 1 month to 8 years
after the patients’ hospitalisation. The hypothesis was that remote,
retroactive intercessory prayer shortens hospital stay and lessens
mortality
after trauma.
Our medical centre’s institutional review board approved this study.
All patients recorded in the trauma registry of a Level 1 trauma
centre
between January 1996 and May 2004 were included in this study. The trauma
centre serves a metropolitan area of greater than 3 million inhabitants
and
has approximately 4,000 trauma admissions per year.
In June 2004 the patients were stratified into subgroups using
bifurcation
values for each of four variables known to have a strong impact on
mortality
after trauma. The variables were age, systolic blood pressure on arrival,
Glasgow Coma Score on arrival, and Injury Severity Score. Their lower
grouping values were age less than or equal to 55 years, systolic blood
pressure on arrival less than or equal to 100mm Hg, Injury Severity Score
less
than or equal to 8, and Glasgow Coma Score on arrival less than or equal
to
8. Within each subgroup, a block randomization into two groups was
performed using a pseudo-random number generator.
Once randomisation was complete the two groups of patients’ names
were
printed out, using the first name and last initial. A coin toss was used
to
designate the study group. The list of names for the study group was given
to
an intercessor who prayed for the health and well being of the people in
the
study group.
Two outcome variables were measured: length of hospital stay, and in-
hospital mortality.
The Mann-Whitney test was used to compare the study and control
groups to
verify they were comparable with regard to age, Glasgow Coma Score on
arrival, systolic blood pressure on arrival, and Injury Severity Score.
For these
variables, a 2-sided analysis was done based on a null hypothesis of no
difference between the two groups.
A 1-sided Mann-Whitney test was used to compare the study and control
groups’ hospital lengths of stay. The test was based on a null hypothesis
that
the stays of members of the control group tended to be shorter.
Fisher’s exact test was used to compare the study and control group’s
mortality rates. This test was 1-sided, where the null hypothesis is that
the
mortality rate of the control group does not exceed that of the prayer
group.
Our sample size of over 33,000 patients was sufficient to detect a 1%
difference in mortality rate, if such a difference existed, with 99% power
with
a Type I error of .05.
Of the 33473 trauma registry patients, 16736 were randomized to the
control
group and 16737 were randomized to the study group. There were no
significant differences between the groups with regard to age, systolic
blood
pressure on arrival, Glasgow Coma Score on arrival, and Injury Severity
Score.
Median length of stay was 2 days for the control group (range 0 to
375 days)
and 2 days for the study group (range 0 to 730 days). The p-value for the
associated test of hypothesis was .27. In-hospital mortality for the
control
group was 4.6% and 4.6% for the study group (p=.44).
Our study’s findings conflict with those reported by Leibovici (1)
who showed
that remote retroactive intercessory prayer was beneficial. It may be
argued
that the outcomes we studied – mortality and hospital stay – were not
appropriate measures to assess improvements in “well being”. It is
arguable
that, in some cases, death may end suffering, and so bring a form of
benefit
to a patient. Similarly, a shorter hospital stay may not necessarily
reflect
improved health or well-being. Additionally, we did not assess the long-
term
health outcome or the patient’s sense of well-being. It is possible there
were
differences in the two groups that we did not assess. Nor did our study
make
an effort to control for prayer that may have been said on the patients’
behalf
by other people, such as family members or friends. However, our aim in
this
study was to assess objective and measurable outcomes that are important
to
clinicians. Our thought (and perhaps it is a biased one) was that survival
was
better than death, and that a shorter hospital stay was better than a long
one.
Although we cannot claim we have shown remote retroactive intercessory
prayer provides no benefit to trauma patients, we think we can claim to
have
shown it is quite unlikely that it affects in-hospital mortality or length
of stay.
This study does not prove that intercessory prayer has no benefits. There
are
several compelling studies which appear to show the benefits of prayer,
and
we feel further research in this area is warranted.
1. Leibovici L. Effects of remote, retroactive intercessory prayer on
outcomes
in patients with bloodstream infection: randomised controlled trial. BMJ
2001,
323:1450-1
Competing interests:
None declared
Competing interests: No competing interests