This is Journal Watch, a medical-literature
survey produced by the Massachusetts Medical
Society.
  Twice a week, our physician-editors summarize
important new clinical research from a group
of 25 journals. Twice a month, the summaries
get compiled into newsletter form and mailed
to subscribers.
  If you'd like to subscribe to the newsletter,
which costs $US79 a year, please call
800-843-6356; international orders:
(49) 30 335 8006.
  To comment via e-mail, please contact us at

  Contents copyright 1994, Mass. Medical Society.

Journal Watch Summaries for June 7, 1994

DETECTING EARLY {*filter*}-CANCER RECURRENCE: DOES IT
HELP?.
JAMA 1994 May 25; 271:1587-92.

PATIENT SATISFACTION WITH MANAGED CARE.
JAMA 1994 May 25; 271:1579-86.

{*filter*} CONSUMPTION IN WOMEN OF CHILDBEARING AGE.
MMWR 1994 May 13; 43:328-9, 335.

SURFACTANT LOWERS MORTALITY AND COSTS IN NEWBORNS.
N Engl J Med 1994 May 26; 330:1476-80.

STRICT PRENATAL DIABETES CONTROL REDUCES NEONATAL
HYPOCALCEMIA.
Obstet Gynecol 1994 Jun; 83:918-22.

DETECTING EARLY {*filter*}-CANCER RECURRENCE: DOES IT
HELP?.
   After surgery for {*filter*} cancer, many women undergo
frequent diagnostic tests aimed at early detection of
recurrence in addition to their annual mammograms. Two
multicenter randomized trials in Italy measured the value of
such testing.
   In the first trial, 1320 women with stage I, II, or III
primary {*filter*} cancer were assigned to intensive
surveillance with laboratory and radiologic testing or to a
control group in which tests were used only as clinically
indicated.
   After a median follow-up of 71 months, survival in the
intensive testing group was similar to that in controls (20
vs. 18 percent). Patients who had intensive testing were
more likely to be asymptomatic at the time of cancer
recurrence (31 vs. 21 percent), but patterns of recurrence
and rates of distant-metastasis-free survival were similar.
Quality-of-life measures did not differ between the two
groups, but more than 70 percent of all respondents said
that they wanted to see their physician often and undergo
diagnostic tests even if asymptomatic.
   The second trial assigned 1243 women to intensive follow-
up with radiologic testing or non-intensive follow-up. The
intensively followed group had increased detection of
intrathoracic and bone metastases (112 vs. 71 cases), but
there was no difference in five-year overall survival, the
main study endpoint.
   Comment: These data indicate that frequent radiologic and
lab testing for recurrent {*filter*} cancer does not improve
patient outcome, and that these tests should not be used
routinely. Their use may be strongly related to patients'
psychological needs. --TH Lee.
   Citation: The GIVIO Investigators. Impact of follow-up testing
on survival and health-related quality of life in {*filter*} cancer
patients: a multicenter randomized controlled trial. JAMA 1994
May 25; 271:1587-92.
   Citation: Del Turco MR; et al. Intensive diagnostic follow-up
after treatment of primary {*filter*} cancer: a randomized trial.
JAMA 1994 May 25; 271:1593-7.

PATIENT SATISFACTION WITH MANAGED CARE.
   Now that changes in the health care system are driving
people into managed care plans, patients and physicians are
expressing concern about the impact of this trend on
quality. This longitudinal study of 1208 {*filter*}s with chronic
disease in Boston, Chicago, and Los Angeles examined how
reimbur{*filter*}t structure affects several aspects of primary
care.
   Data were collected by questionnaire from 1986 to 1990.
Patients were in one of three types of plans: fee-for-
service (FFS), a prepaid independent practice association
(IPA), or a prepaid health maintenance organization. FFS
care ranked highest in logistical accessibility and in
continuity of care. HMOs ranked first in financial
accessibility and coordination of care, but were worst in
comprehensiveness of care. Patients in the two prepaid plans
rated providers lower on accountability (particularly
interpersonal manner) than did FFS patients. The variability
in patient opinions about the continuity and coordination of
care in prepaid plans was more than double that among FFS
patients.
   Comment: These data show important differences in the
perceived quality of primary care under different systems.
Prepaid and FFS systems each have both strengths and
weaknesses. An important goal of health care reform should
be to improve both types of care. --TH Lee.
   Citation: Safran DG; et al. Primary care performance in fee-
for-service and prepaid health care systems: results from the
Medical Outcomes Study. JAMA 1994 May 25; 271:1579-86.

{*filter*} CONSUMPTION IN WOMEN OF CHILDBEARING AGE.
   Both daily excessive {*filter*} consumption and binge
drinking during pregnancy are associated with adverse
perinatal outcomes. This report describes self-reported
{*filter*} consumption for over 26,500 U.S. women of
childbearing age who took part in the Behavi{*filter*}Risk Factor
Surveillance System survey.
   Fifty percent of respondents said that they had not
consumed {*filter*} during the past month, 45 percent reported
light drinking (30 or fewer drinks), 3 percent reported
moderate drinking (31 to 59), and 2 percent reported heavy
drinking (60 or more). The 21 percent of women who reported
binges (five or more drinks on one occasion) had
substantially higher regular consumption rates. Women in
northern states had the highest rate of moderate, heavy, or
binge drinking. Of the 1067 women who reported being
pregnant at the interview, 14 percent reported at least
light drinking during the preceding month, and 1.3 percent
reported binge drinking.
   Comment: Although excessive {*filter*} consumption during
pregnancy is known to be harmful, the evidence remains
equivocal for light drinking. Should all women of
childbearing age (pregnant or not) be cautious about
consuming {*filter*}? I am wary of alarming women who have
imbibed occasionally before they knew they were pregnant,
since the evidence is not sufficient to warrant concern. It
is heartening to see that pregnant women do drink much less
than nonpregnant women. --DM Berwick.
Frequent {*filter*} consumption among women of childbearing age --
Behavi{*filter*}Risk Factor Surveillance System, 1991. MMWR 1994 May
13; 43:328-9, 335.

SURFACTANT LOWERS MORTALITY AND COSTS IN NEWBORNS.
   In the late 1980s, clinical trials demonstrated that
surfactant improves the outcome of very-low-birth-weight
infants. This study examined the effects of surfactant in 14
U.S. tertiary perinatal units after its introduction into
clinical practice.
   Researchers analyzed clinical and financial data from
5629 infants weighing 500 to 1500 grams; 3922 were born
before and 1707 after the release of surfactant for general
use. No other therapeutic innovations were introduced during
the transition period.
   Overall mortality decreased significantly from 24 percent
before to 20 percent after the release of surfactant. This
benefit was confined to infants weighing 500 to 1250 grams.
After adjustment for sex, race, and birth weight, the
relative reduction in mortality was 30 percent overall and
40 percent among infants with bronchopulmonary dysplasia.
Hospital charges decreased 10 percent for survivors and 31
percent for infants who died.
   Comment: The benefits of surfactant shown in clinical
trials have held up and remain demonstrable in routine
clinical practice. --AS Brett.
   Citation: Schwartz RM; et al. Effect of surfactant on
morbidity, mortality, and resource use in newborn infants
weighing 500 to 1500 g. N Engl J Med 1994 May 26; 330:1476-80.

STRICT PRENATAL DIABETES CONTROL REDUCES NEONATAL
HYPOCALCEMIA.
   Thirty to fifty percent of infants born to insulin-
dependent diabetic mothers have hypocalcemia. This
randomized trial asked whether strict management of diabetes
during pregnancy affects the incidence of neonatal
hypocalcemia.
   Researchers randomized 137 insulin-dependent diabetic
women who were less than nine weeks pregnant to strict
glycemic control ({*filter*} glucose goals of less than 80 mg/dl
fasting and 120 mg/dl 1.5 hours after eating) or customary
glycemic control (less than 100 fasting and 140 mg/dl
postprandial). All women received insulin twice daily. The
women assigned to strict control were hospitalized initially
to achieve glucose goals and followed weekly thereafter;
those assigned to customary control were hospitalized after
one week if glucose goals were not achieved and followed
biweekly thereafter.
   Infants born to mothers in the strict control group had
significantly less hypocalcemia (18 vs. 32 percent) and
hypomagnesemia (10 vs. 15 percent) than infants of mothers
in the customary control group. Maternal glycosylated
hemoglobin levels did not differ between the two groups.
   Comment: This interesting trial suggests that a regimen
aimed at strict control of maternal diabetes and close
follow-up may help prevent neonatal hypocalcemia. The
mechanism and clinical significance of these findings remain
uncertain. --CD Mulrow.
   Citation: Demarini S; et al. Impact of metabolic control of
diabetes during pregnancy on neonatal hypocalcemia: a randomized
study. Obstet Gynecol 1994 Jun; 83:918-22.

EXTRAPYRAMIDAL SIGNS IN NURSING HOME RESIDENTS.
   The Omnibus Budget Reconciliation Act in 1991 mandates
that the use of antipsychotic medications in long-term care
facilities be linked to specific criteria. Antipsychotic use
has since decreased, but many nursing home residents remain
on these {*filter*}. This study in 12 Massachusetts nursing homes
examined antipsychotic use and its effects on extrapyramidal
function.
   An independent examiner who was unaware of medication use
assessed 251 elderly nursing home residents who were taking
psychoactive {*filter*}, and found parkinsonian signs in 127.
Overall, 111 residents were taking antipsychotics, most
often haloperidol and thioridazine. Significant predictors
of parkinsonian signs were dementia (Mini-Mental State
Examination score below 20), use of 1 mg or more daily of
haloperidol, and use of other antipsychotic {*filter*} (including
low-potency agents such as thioridazine) in doses of 50 or
more chlorpromazine-equivalent milligrams daily.
Benzodiazepines, antidepressants, and low-dose
antipsychotics were not associated with parkinsonian signs.
   Comment: These findings highlight the need to limit the
use of antipsychotics, particularly in high doses, in
elderly nursing home residents. Randomized trials are sorely
needed to establish the benefit-to-risk ratios of these
{*filter*} among frail elders. --CD Mulrow.
   Citation: Avorn J; et al. Clinical assessment of
extrapyramidal signs in nursing home residents given
antipsychotic medication. Arch Intern Med 1994 May 23; 154:1113-
7.