Hoffman JR  Reynolds S
Comparison of nitroglycerin, morphine and furosemide in treatment of
  presumed pre-hospital pulmonary edema.
Chest (1987 Oct) 92(4):586-93

We compared four treatment protocols in 57 patients with presumed pre-
  hospital pulmonary edema. Group A patients were treated with
  nitroglycerin and furosemide, group B patients with morphine sulfate
  and furosemide, group C with all three agents, and group D with
  nitroglycerin and morphine, but without furosemide. Twenty-three
  percent of our patients were ultimately found not to have pulmonary
  edema, with pneumonia and/or exacerbations of chronic lung disease
  the most frequent alternate diagnoses. Group A patients had
  significantly greater improvement, both subjectively and objectively,
  than group B patients, a substantial number of whom failed to
  improve, or even worsened. There was no evident synergistic effect of
  any of the drugs, and some indication that furosemide might have
  caused clinically important problems with fluid and electrolyte
  management in some of the patients. Our data suggest that
  nitroglycerin is beneficial in the management of presumed pre-
  hospital pulmonary edema, while morphine and furosemide may not add
  anything to its efficacy, and may be potentially deleterious in some
  of these patients.

Wuerz RC  Meador SA
Effects of prehospital medications on mortality and length of stay in
Ann Emerg Med (1992 Jun) 21(6):669-74

HYPOTHESIS: Prehospital medications for congestive heart failure
  should affect hospital outcomes (survival and length of stay). STUDY
  DESIGN: In a retrospective case series, hospital outcomes were
  compared for patients treated with prehospital nitroglycerin,
  furosemide, and/or morphine (252) versus those given no medications
  (241). SETTING: A rural/suburban emergency medical services system
  (population 140,000) served by three paramedic units. PARTICIPANTS:
  Four hundred ninety-three consecutive cases of congestive heart
  failure or pulmonary edema were identified by hospital discharge
  diagnosis from a data base of 8,315 paramedic transports with known
  outcome. INTERVENTIONS: Oxygen was given by protocol to 489 patients.
  Other medications were given by order of on-line physician medical
  command. RESULTS: Overall mortality was 10.9% (54 of 493). Treated
  and untreated patients were comparable in age, sex, cardiac rhythms,
  prior use of cardiac medications, and response and scene times;
  mortality was reduced in treated versus untreated patients (odds
  ratio for improved survival, 2.51; 95% confidence interval, 1.37 to
  4.55; P less than .01). Positive treatment effect was greatest for 58
  nonhypotensive, critical patients (odds ratio for survival, 10.25; P
  less than .01). No single drug combination was unique in terms of
  treatment benefit. Patients treated in the field received medications
  36 minutes earlier than patients first treated in the emergency
  department. No survival benefit was evidence for noncritical,
  nonhypotensive patients, and patients with final diagnoses of asthma,
  chronic obstructive pulmonary disease, pneumonia, or bronchitis had a
  higher than expected mortality if erroneously treated for congestive
  heart failure. Differences in hospital length of stay were not
  significant for any group. CONCLUSION: Prehospital medications
  improve survival in congestive heart failure, especially in critical
  patients. More than one combination of medications seems effective,
  and early treatment is associated with improved survival. However,
  these medications appear to increase mortality in patients
  misdiagnosed in the field. Factors used in paramedica and medical
  command assessments require further study.

REFERENCE: "Is Ambulance Transport Time with Lights and Siren faster than
that without?" Hunt, R.C., et al, Ann Emerg Med 25(4):507, April, 1995
ABSTRACT:
BACKGROUND:  It has been reported that up to 12,000 emergency medical
vehicle crashes occur each year in the U.S. and Canada as a direct result
of the use of warning lights and sirens (L&S). Despite the common use of
L&S in ambulance transport, this practice has not been demonstrated to be
life-saving.  METHODS:  This prospective study, from East Carolina
University School of Medicine in Greenville, NC, compared ambulance
transport times with and without L&S.  L&S transports of 50 patients were
timed and mapped by an observer, and the runs subsequently duplicated
without L&S at the same time of day and same day of the week (with
paramedic drivers instructed to obey speed limits and traffic laws).  All
transports were to an ED in a city with a population of 46,000, and
involved distances of less than nine miles.  RESULTS:  Overall, 76% of
transports were faster with L&S, which averaged 44 seconds faster than
non-L&S transports (p=0.0001).  The differences between L&S and non-L&S
transport times ranged between more than five minutes faster to almost
three minutes slower with L&S.  However, the largest improvement in
transport time was an outlier, as the second fastest L&S transport was less
than three minutes faster than its corresponding non-L&S transport.
CONCLUSIONS:  The authors suggest that, except in rare circumstances, the
time savings of less than one minute achieved with L&S does not warrant use
of L&S during prehospital ambulance transport.  The authors suggest that
each system develop guidelines that rationally balance time saved with L&S
transport against the associated risks of this practice.  11 references