- Email: [email protected]
Outcome of patients who refused out-of-hospital medical assistance
Outcome of Patients Who Refused Out-of-Hospital Medical Assistance JONATHAN L. BURSTEIN, MD,* MARK C. HENRY, MD,* JEANNE ALICANDRO, MD,* DAVID GENTILE, EMT-D,I HENRY C. THODE, JR, PHD,* JUDD E. HOLLANDER, MD* Activation of the emergency medical services (EMS) system does not always result in transport of a patient to the hospital. This study assessed the outcomes of patients who refused medical assistance in the field, to determine if refusal of medical assistance (RMA) is associated with poor outcomes. Four high-volume suburban volunteer ambulance corps participated in the study. Consecutive patients who refused medical assistance were prospectively enrolled, Medical and identifying data were collected for each patient. Telephone follow-up was conducted to determine the patient's condition and if the patient sought further care after RMA. Primary endpoints were whether the patient sought further care, was admitted to a hospital, or died subsequent to RMA. Follow-up was successfully obtained for 199 of 321 patients enrolled (62%). Of these 199 patients, 95 (48%) sought further medical care within 1 week for the same complaint, with 13 being admitted to the hospital. Six of the 13 admitted patients had chief complaints of a cardiac or respiratory nature. One patient died during hospital admission. Even if none of the patients lost to follow-up had sought further care, a substantial number of patients who refuse out-of-hospital medical assistance seek further care. Some of these patients require hospital admission, especially those with cardiac or respiratory complaints, Efforts to minimize RMA should be especially focused on patients with such complaints. (Am J Emerg Med 1996;14:2326, Copyright © 1996 by W.B. Saunders Company) Activation of the emergency medical services system (EMS) usually results in transport of an identified patient to a medical facility for further care. In some systems, however, as many as 25% to 70% of all EMS system activations will result in no transport. 1At times, the reason for nontransport is refusal of medical assistance (RMA) by the patient. Various reasons may account for patients' R M A : the identified patient m a y decide that he or she no longer wants to go to the hospital for further care, or the patient was not the one who activated the EMS system and saw no need for help in the first place. Intoxication with alcohol or drugs, and/or altered mental status, m a y also result in RMA. 2-7 Some patients who refuse medical assistance have been shown to be at risk for poor medical outcomes. 1,5,8-~°,11 The
From the *Department of Emergency Medicine, University Medical Center, State University of New York, Stony Brook, and the ]Division of Emergency Medical Services, Suffolk County Department of Health Services, Yaphank, NY. Received January 25, 1995, returned February 21, 1995; revision received March 3, 1995, accepted March 9, 1995. Presented at the 1995 American College of Emergency Physicians Research Forum. Address reprint requests to Dr Burstein, Department of Emergency Medicine, University Medical Center, Level 4, Room 515, State University of New York, Stony Brook, New York 11794- 7400. Key Words: Emergency medical services, follow-up, medical control, liability, refusal of transport, patient transport. Copyright © 1996 by W.B. Saunders Company 0735-6757/96/1401-000755.00/0
patient's refusal may put either the EMS providers or the system at substantial legal risk, 2,5,6,8,12-14 especially if the patient has a poor outcome. Only a few studies have looked at medical outcomes after RMA. 1,2,n They were hindered by small sample size and poor follow-up. This study assessed the outcomes of patients who refused medical assistance to determine whether the failure to transport these patients to the hospital results in adverse outcomes.
METHODS
Study Design This was a prospective observational cohort study of the frequency of further physician contact, hospitalization, or death for patients who refused out-of-hospital medical assistance. The study was conducted over a 3-month period in 1994.
Emergency Medical Services System Suffolk County, New York is a suburban/rural county on eastern Long Island. The county covers 932 square miles and has a population of 1.4 million. The Suffolk County Depaitment of Health Services administers the EMS system, which provides both basic and advanced life support care. Fire, rescue, and emergency medical services throughout the region are provided on an allvolunteer basis by 111 fire departments, 63 of whom provide ambulance service, and 31 volunteer ambulance corps. The emergency ambulance services in the county respond to more than 77,000 calls each year. The Department of Emergency Medicine of the University Medical Center at Stony Brook serves as the sole on-line medical control facility for all 12 ambulance receiving hospitals in Suffolk County, Providers performing advanced life support (ALS) services contact medical control where care is directed by one of 22 emergency physicians certified as base station medical control physicians. Basic life supports (BLS) providers are not required to contact the medical control facility. Console physicians are emergency medicine faculty and senior residents who have completed a full-day seminar in medical control operation and ALS protocol review, and have passed a written examination. Daily calls are further reviewed by the physician directing EMS continuous quality improvement. New York State policy requires that a patient must be alert, oriented, and understand the consequences before refusing medical care. Minors, incompetent patients, and intoxicated patients are not permitted to refuse medical assistance. In Suffolk County, however, there is no mechanism to enforce transport against a patient's wishes except via the Suffolk County Police Department. The police often respond to ambulance calls, and will respond if requested by an EMS crew or medical control physician. The police, however, make an independent determination of the need to place the patient in protective custody for transport (usually done only if they deem the patient to be an immediate threat to self or 23
24
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
others). If the police refuse to place the patient in custody, the EMS personnel have no legal or practical mechanism to enforce transport.
Population and Setting Four volunteer corps with very high call volumes were solicited and agreed to participate in the study. These corps consisted of both BLS and ALS personnel. Consecutive patients who refused medical assistance were enrolled. Patients were excluded if the ambulance had not been called specifically for the patient, such as an ambulance response to a routine police call to a motor vehicle accident scene where there were no injured patients. Patients who initially attempted RMA but were then transported to the hospital were not enrolled. The study was approved by the Suffolk County EMS Medical Director. This study was conducted under a contractual agreement between Suffolk County and University Medical Center of the State University of New York at Stony Brook to monitor quality improvement in Suffolk County EMS. It was exempt from review by the University Medical Center institutional review board.
TABLE2. Chief Complaints of Patients Refusing Medical Assistance According to Whether Follow-Up Was Obtained
Chief Complaint
Follow-Up (n = 199)
No Follow-Up (n = 122)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
1% 15% 1% 4% 9% 6% 11% 2% 2% 2% 47%
3% 18% 3% 3% 9% 9% 5% 3% 2% 9% 36%
NOTE: Numbers in each box expressed as percentage of total patients in that column.
Data Analysis Protocol and Experimental Design Out-of-hospital personnel prospectively enrolled patients who refused medical assistance and transport to the hospital. The EMS crews were required to complete an identifying card on each RMA patient, including demographic data, telephone number, address, and chief complaint. The chief complaint was prospectively classified into one of 11 categories by the study investigators based on the information available from the prehospital care report (PCR). These categories are listed in Table 1. Follow-up was attempted by telephone for a period of 2 weeks after the RMA. Telephone numbers were obtained from the identifying cards, telephone directories, directory assistance, or Coles reverse telephone directory. Patients were telephoned up to three times over the course of the study between the hours of 0900 and 2100. Messages and a callback number were left, if possible, for those patients who were not at that number at the time of the call. For those patients without a telephone, the telephone numbers of the two neighbors geographically closest were obtained from the reverse directory. These numbers were called, and messages and a call-back number were left for the patient. When the patient or family was contacted, data collected included the patient's further attempts to seek medical care for the same condition that had prompted the initial call for help, whether through the EMS system, hospital emergency departments (EDs) or physicians, and the time course of such attempts. Hospital admissions and deaths were noted.
TABLE1. Chief Complaints of All Patients Refusing Medical Assistance
Chief Complaint
No. of Patients (n = 321)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
5 52 6 12 29 23 28 7 5 15 139
The outcome parameters were whether patients who refused medical assistance later sought medical care for the same condition through any route, whether patients were admitted to the hospital, or whether patients died. Data are presented as percentages of patients in a particular outcome category. Chi-square or Fisher exact tests were used to compare outcome category with chief complaint. Statistical significance for all comparisons was defined as a two- tailed alpha value <0.05 for all tests.
RESULTS During the course of the study, 361 patients attempted R M A from the four involved corps. Forty patients were eventually transported to the hospital by EMS personnel and therefore excluded from the study; one third of these were convinced to be transported after contact with the base station physician. Physician contact resulted in an increased rate of transport; these data have been reported elsewhere. 1° The remaining 321 patients (or parents/guardians) refused assistance despite attempts at persuasion by the EMS personnel or medical control physicians and constituted the study group. The average age of these patients was 39 years (range 0 to 94 years). The patients' chief complaints are shown in Table 1. Follow-up was obtained for 199 patients (62%) by telephone contact. The remaining 122 patients (38%) could not be contacted. The reasons for failure to contact were as follows: 53 patients had no phone and could not be reached through their neighbors, 34 did not return repeated phone messages, 17 had insufficient demographic data to be identified for follow-up, and 18 for other reasons. Comparison of the patients lost to follow-up with those who were followed up is shown in Table 2. Patients with a chief complaint of intoxication were more likely to be lost to follow-up, and patients with neurological or traumatic complaints were less likely to be lost to follow-up (P < .04). The average age of those who were followed up was greater than that of those lost to follow-up (42 years vs 34 years; P < .001). Of the 199 patients who were followed up, 95 (48%) sought further medical care; 104 (52%) had not sought and did not desire further medical care. The 95 patients who
BURSTEIN ET AL • RMA OUTCOME
25
sought further care represent 30% of all patients who refused medical assistance by EMS. All those who sought further care did so within 1 week. The distribution of chief complaints in these two groups is shown in Table 3. No significant difference was found between the two groups in the type or frequency of chief complaints (P = .39), although patients with cardiac/respiratory complaints tended to be more likely to seek care. Of those who sought further care, 51 were ultimately seen in an ED; 47 (49%) visited their private physician within 1 week of their RMA (3 of these were sent to an ED by their physician). The other 48 patients (51%) presented to an ED directly for care, 7 of them by again accessing EMS. Thirteen patients (25%) were admitted to the hospital, and one died in hospital. Six of the 13 admitted patients had cardiac or respiratory chief complaints. The admitted patients' discharge diagnoses are shown in Table 4.
DISCUSSION The problem of patients who refuse medical care in the out-of-hospital setting is ubiquitous in EMS systems. 1-4.6-10.12 These patients may suffer poor outcomes following RMA, and this may also lead to litigation against the EMS system. 5,6,8,9,12J3 Morgan et all4 found that 34% of EMS patient care negligence cases filed were due to failure or delay of transportation. Outcome studies in this patient population are difficult to conduct because of the nature of the out-of- hospital environment. Nevertheless, several small analyses have been conducted. Zachariah et al a reported a retrospective analysis of 158 cases of out-of-hospital RMA. By telephone follow-up, they were able to determine the outcome of 59% of these patients, and found that 60 patients sought further care, with 15 being admitted to a hospital, 2 of whom died. In the system they analyzed, EMS crews were able to refuse transport in certain cases, and 11 of the 15 admissions were in this group. They concluded that patients not transported were at risk of serious or fatal outcomes, and that although causation had not been established, EMS refusal to transport was associated with higher risk to the patient. They noted that their
TABLE3. Chief Complaints of Patients in Whom Follow-Up Was Obtained According to Whether They Sought Additional Medical Care Chief Complaint
Care Obtained (n = 95)
No Care Obtained (n = 104)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
1% 20% 0% 3% 8% 5% 10% 3% 1% 1% 48%
1% 10% 2% 5% 10% 7% 13% 0% 2% 3% 47%
NOTE: Numbers in each box expressed as percentage of total patients in that column.
TABLE4. Chief Complaint and Ultimate Hospital Discharge Diagnosis of Patients Who Initially Refused Medical Care But Later Were Admitted to Hospital Case ID
Chief Complaint
28 39
Cardiac/respiratory Trauma
54
Miscellaneous
93 113 147
Miscellaneous (weakness/nausea) Cardiac/respiratory Cardiac/respiratory
188 244 283 289 375 419 436
Cardiac/respiratory Obstetric/gynecological Cardiac/respiratory Cardiac/respiratory Trauma Trauma Diabetic-related
Discharge Diagnosis Myocardial infarction Diabetes mellitus and extremity trauma Gastroenteritis, dehydration Digoxin toxicity (level = 7.5) Suicidal ideation Congestive heart failure (died) Multiple sclerosis Normal vaginal delivery Myocardial ischemia Congestive heart failure Myocardial ischemia Rheumatoid arthritis Congestive heart failure
findings indicated the need for further, prospective study of the risks of nontransport. Sucov et al 2 retrospectively found 188 cases of patients who refused medical assistance in the out-of-hospital setting. No patients were refused transport by the EMS crews. By telephone follow-up, they were able to determine outcomes in 50% of these patients. Of these patients, 31 contacted a physician and 6 were admitted to a hospital. They note that this number probably underestimates the actual number admitted because of the relatively low follow-up achieved. They call for "additional studies to provide a better description of the natural outcome" of patients after RMA. Cone et aP 1 retrospectively analyzed 81 cases of patientinitiated refusal of prehospital care, obtaining a follow-up rate of 67% (54 patients). They report that 32% of these patients sought further care after RMA, with 7 (13%) admitted to the hospital. There were no deaths in their study group. They noted that on-line medical control appeared to improve documentation in cases of RMA, and called for research into the effects of physician contact on RMA. Alicandro et ali0 found that physician contact improves the transport rate of patients who attempt RMA. Our study is the first prospective study to address the issue of the natural outcome of patients after RMA. Since our system does not permit EMS refusal of transport, our study population is comparable to that of Sucov et al. 2 We found that 48% of the patients we followed up sought further care after RMA. This is comparable to the previously reported 32% to 64% from the retrospective studies. 1,2,~1 Thirteen patients in our study were admitted to the hospital; this 7% admission rate is somewhat lower than the other studies reported. It is possible that our study design was able to find more nonadmitted patients for follow-up, or that greater effort was made by out-of-hospital providers to transport ill patients in this prospective study. Although the number of patients is too low to be definitive, we note that 6 of 13 admitted patients had cardiac or respiratory chief corn-
26
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
plaints, suggesting the possibility that these patients are at increased likelihood of requiring admission after RMA. Our study has several weaknesses. We report on 321 patients with a 20% improvement on follow-up (50% to 62%) compared with the study of Sucov et al,2 but we note that 38% of patients in our study were still lost to follow-up. This group included patients with serious chief complaints, such as altered mental status, cardiac or respiratory problems, and trauma. It is possible that significant outcomes were missed in this group lost to follow-up. Even if we make the unlikely assumption that all of the patients lost to follow-up did not seek further medical care, 30% of our total patient cohort that refused medical assistance would still have sought further care. Our study also has an observer bias, because the EMS crews knew we were analyzing RMA and may have striven to transport ill patients. This would have resulted in an underestimate of the frequency of adverse outcomes in patients who refused medical assistance. Our population is also specifically suburban/rural, and results may not generalize to urban populations. We also could not determine whether delay in obtaining care resulted in any increase in morbidity. Patients who interact with the EMS system may require transport to a hospital for admission, may receive interventions in the ED after transport to prevent admission, or may not require transport at all. It is possible that some patients may be better served by RMA followed by care given by an office-based physician. Our results, however, indicate that patients who refuse medical assistance may not always be able to make the appropriate distinctions of medical necessity. Detailed guidelines need to be developed to determine which patients may be safely allowed RMA as opposed to those in whom great effort should be made to ensure transport to a hospital. Until such criteria are developed, efforts should be made to convince all patients who attempt RMA to be transported to the hospital. Intervention by an on-line medical control physician has been shown to increase transport rates of patients who initially desire RMA,10 and would therefore be an option for EMS systems. Further study to address these issues is warranted.
The authors acknowledge the participation of the following volunteer corps and EMS personnel: Bayshore-Brightwaters Rescue Ambulance: Steve Engelberg, EMT-CC, Tess Melquenn, EMT-P, and Ted Suida, EMT-P; Brentwood Legion Ambulance: Juan Espinosa, EMTCC, Susan O'Malley, EMT-CC; Central Islip- Hauppauge Volunteer Ambulance: Dorothy Norman, EMT-CC, Joseph Ortiz, EMT-D, Joseph Tricario, EMT-D, and Meredith Whipple, EMT-D; Commack Volunteer Ambulance: Nancy Wichtendahl, RN, EMT-CC; State University of New York at Stony Brook: Robert Delagi, EMT-P.
REFERENCES 1. Zachariah BS, Bryan D, Pepe PE, Griffin M: Follow-up and outcome of patients who decline or are denied transport by EMS, Prehospital Disaster Medicine 1992;7:359-363 2. Sucov A, Verdile VP, Garettson D, Pads PM: The outcome of patients refusing prehospital transportation. Prehospital Disaster Medicine 1992;7:365-371 3. Holroyd B, Shalit M, Kallsen G, et al: Prehospital patients refusing care. Ann Emerg Med 1988;17:957-963 4. Stark G, Hedges JR, Neely K, Norton R: Patients who initially refuse prehospital evaluation and/or therapy. Am J Emerg Med 1990;8:509-511 5. American College of Emergency Physicians: Continuous Quality Improvement in EMS. Dallas, TX, ACEP, 1992, pp 298- 303 6. Mottley L: Refusal of medical assistance in the field. In Kuehl AE (ed): Prehospital Systems and Medical Oversight. St. Louis, MO, National Association of EMS Physicians and CV Mosby, 1994, pp 375-380 7. Selden BS, Schnitzer PG, Nolan FX, Veronesi JF: The "nopatient" run: 2698 patients evaluated but not transported by paramedics. Prehospital Disaster Medicine 1991 ;6:135-142 8. Selden BS, Schnitzer PG, Nolan FX: Medicolegal documentation of prehospital triage. Ann Emerg Med 1990;19:547- 551 9. Goldberg RJ, Zautcke JL, Konigsberg MD, et al: A review of prehospital care litigation in a large metropolitan EMS system. Ann Emerg Med 1990; 19:557-561 10. Alicandro JM, Hollander JE, Henry MC, et al: Impact of interventions for patients refusing EMS transport. Academic Emerg Med 1995;2:480-485 11. Cone DC, Kim DT, Davidson SJ: Patient-initiated refusals of prehospital care: Ambulance call report documentation, patient outcome, and on-line medical command. Prehospital Disaster Medicine 1995; 10:3-9 12. Soler JM, Montes MF, Egol AB, et al: The ten year malpractice experience of a large urban EMS system. Ann Emerg Med 1985;14: 982-985 13. Goldstein A: EMS and the Law: A legal handbook for EMS personnel. Bowie, MD, Robert J. Brady Co, 1983, pp 69-82 14. Morgan DL, Wainscott MP, Knowles HC: Emergency medical services liability litigation in the United States: 1987 to 1992. Prehospital Disaster Medicine 1994;9:214-221
From the *Department of Emergency Medicine, University Medical Center, State University of New York, Stony Brook, and the ]Division of Emergency Medical Services, Suffolk County Department of Health Services, Yaphank, NY. Received January 25, 1995, returned February 21, 1995; revision received March 3, 1995, accepted March 9, 1995. Presented at the 1995 American College of Emergency Physicians Research Forum. Address reprint requests to Dr Burstein, Department of Emergency Medicine, University Medical Center, Level 4, Room 515, State University of New York, Stony Brook, New York 11794- 7400. Key Words: Emergency medical services, follow-up, medical control, liability, refusal of transport, patient transport. Copyright © 1996 by W.B. Saunders Company 0735-6757/96/1401-000755.00/0
patient's refusal may put either the EMS providers or the system at substantial legal risk, 2,5,6,8,12-14 especially if the patient has a poor outcome. Only a few studies have looked at medical outcomes after RMA. 1,2,n They were hindered by small sample size and poor follow-up. This study assessed the outcomes of patients who refused medical assistance to determine whether the failure to transport these patients to the hospital results in adverse outcomes.
METHODS
Study Design This was a prospective observational cohort study of the frequency of further physician contact, hospitalization, or death for patients who refused out-of-hospital medical assistance. The study was conducted over a 3-month period in 1994.
Emergency Medical Services System Suffolk County, New York is a suburban/rural county on eastern Long Island. The county covers 932 square miles and has a population of 1.4 million. The Suffolk County Depaitment of Health Services administers the EMS system, which provides both basic and advanced life support care. Fire, rescue, and emergency medical services throughout the region are provided on an allvolunteer basis by 111 fire departments, 63 of whom provide ambulance service, and 31 volunteer ambulance corps. The emergency ambulance services in the county respond to more than 77,000 calls each year. The Department of Emergency Medicine of the University Medical Center at Stony Brook serves as the sole on-line medical control facility for all 12 ambulance receiving hospitals in Suffolk County, Providers performing advanced life support (ALS) services contact medical control where care is directed by one of 22 emergency physicians certified as base station medical control physicians. Basic life supports (BLS) providers are not required to contact the medical control facility. Console physicians are emergency medicine faculty and senior residents who have completed a full-day seminar in medical control operation and ALS protocol review, and have passed a written examination. Daily calls are further reviewed by the physician directing EMS continuous quality improvement. New York State policy requires that a patient must be alert, oriented, and understand the consequences before refusing medical care. Minors, incompetent patients, and intoxicated patients are not permitted to refuse medical assistance. In Suffolk County, however, there is no mechanism to enforce transport against a patient's wishes except via the Suffolk County Police Department. The police often respond to ambulance calls, and will respond if requested by an EMS crew or medical control physician. The police, however, make an independent determination of the need to place the patient in protective custody for transport (usually done only if they deem the patient to be an immediate threat to self or 23
24
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
others). If the police refuse to place the patient in custody, the EMS personnel have no legal or practical mechanism to enforce transport.
Population and Setting Four volunteer corps with very high call volumes were solicited and agreed to participate in the study. These corps consisted of both BLS and ALS personnel. Consecutive patients who refused medical assistance were enrolled. Patients were excluded if the ambulance had not been called specifically for the patient, such as an ambulance response to a routine police call to a motor vehicle accident scene where there were no injured patients. Patients who initially attempted RMA but were then transported to the hospital were not enrolled. The study was approved by the Suffolk County EMS Medical Director. This study was conducted under a contractual agreement between Suffolk County and University Medical Center of the State University of New York at Stony Brook to monitor quality improvement in Suffolk County EMS. It was exempt from review by the University Medical Center institutional review board.
TABLE2. Chief Complaints of Patients Refusing Medical Assistance According to Whether Follow-Up Was Obtained
Chief Complaint
Follow-Up (n = 199)
No Follow-Up (n = 122)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
1% 15% 1% 4% 9% 6% 11% 2% 2% 2% 47%
3% 18% 3% 3% 9% 9% 5% 3% 2% 9% 36%
NOTE: Numbers in each box expressed as percentage of total patients in that column.
Data Analysis Protocol and Experimental Design Out-of-hospital personnel prospectively enrolled patients who refused medical assistance and transport to the hospital. The EMS crews were required to complete an identifying card on each RMA patient, including demographic data, telephone number, address, and chief complaint. The chief complaint was prospectively classified into one of 11 categories by the study investigators based on the information available from the prehospital care report (PCR). These categories are listed in Table 1. Follow-up was attempted by telephone for a period of 2 weeks after the RMA. Telephone numbers were obtained from the identifying cards, telephone directories, directory assistance, or Coles reverse telephone directory. Patients were telephoned up to three times over the course of the study between the hours of 0900 and 2100. Messages and a callback number were left, if possible, for those patients who were not at that number at the time of the call. For those patients without a telephone, the telephone numbers of the two neighbors geographically closest were obtained from the reverse directory. These numbers were called, and messages and a call-back number were left for the patient. When the patient or family was contacted, data collected included the patient's further attempts to seek medical care for the same condition that had prompted the initial call for help, whether through the EMS system, hospital emergency departments (EDs) or physicians, and the time course of such attempts. Hospital admissions and deaths were noted.
TABLE1. Chief Complaints of All Patients Refusing Medical Assistance
Chief Complaint
No. of Patients (n = 321)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
5 52 6 12 29 23 28 7 5 15 139
The outcome parameters were whether patients who refused medical assistance later sought medical care for the same condition through any route, whether patients were admitted to the hospital, or whether patients died. Data are presented as percentages of patients in a particular outcome category. Chi-square or Fisher exact tests were used to compare outcome category with chief complaint. Statistical significance for all comparisons was defined as a two- tailed alpha value <0.05 for all tests.
RESULTS During the course of the study, 361 patients attempted R M A from the four involved corps. Forty patients were eventually transported to the hospital by EMS personnel and therefore excluded from the study; one third of these were convinced to be transported after contact with the base station physician. Physician contact resulted in an increased rate of transport; these data have been reported elsewhere. 1° The remaining 321 patients (or parents/guardians) refused assistance despite attempts at persuasion by the EMS personnel or medical control physicians and constituted the study group. The average age of these patients was 39 years (range 0 to 94 years). The patients' chief complaints are shown in Table 1. Follow-up was obtained for 199 patients (62%) by telephone contact. The remaining 122 patients (38%) could not be contacted. The reasons for failure to contact were as follows: 53 patients had no phone and could not be reached through their neighbors, 34 did not return repeated phone messages, 17 had insufficient demographic data to be identified for follow-up, and 18 for other reasons. Comparison of the patients lost to follow-up with those who were followed up is shown in Table 2. Patients with a chief complaint of intoxication were more likely to be lost to follow-up, and patients with neurological or traumatic complaints were less likely to be lost to follow-up (P < .04). The average age of those who were followed up was greater than that of those lost to follow-up (42 years vs 34 years; P < .001). Of the 199 patients who were followed up, 95 (48%) sought further medical care; 104 (52%) had not sought and did not desire further medical care. The 95 patients who
BURSTEIN ET AL • RMA OUTCOME
25
sought further care represent 30% of all patients who refused medical assistance by EMS. All those who sought further care did so within 1 week. The distribution of chief complaints in these two groups is shown in Table 3. No significant difference was found between the two groups in the type or frequency of chief complaints (P = .39), although patients with cardiac/respiratory complaints tended to be more likely to seek care. Of those who sought further care, 51 were ultimately seen in an ED; 47 (49%) visited their private physician within 1 week of their RMA (3 of these were sent to an ED by their physician). The other 48 patients (51%) presented to an ED directly for care, 7 of them by again accessing EMS. Thirteen patients (25%) were admitted to the hospital, and one died in hospital. Six of the 13 admitted patients had cardiac or respiratory chief complaints. The admitted patients' discharge diagnoses are shown in Table 4.
DISCUSSION The problem of patients who refuse medical care in the out-of-hospital setting is ubiquitous in EMS systems. 1-4.6-10.12 These patients may suffer poor outcomes following RMA, and this may also lead to litigation against the EMS system. 5,6,8,9,12J3 Morgan et all4 found that 34% of EMS patient care negligence cases filed were due to failure or delay of transportation. Outcome studies in this patient population are difficult to conduct because of the nature of the out-of- hospital environment. Nevertheless, several small analyses have been conducted. Zachariah et al a reported a retrospective analysis of 158 cases of out-of-hospital RMA. By telephone follow-up, they were able to determine the outcome of 59% of these patients, and found that 60 patients sought further care, with 15 being admitted to a hospital, 2 of whom died. In the system they analyzed, EMS crews were able to refuse transport in certain cases, and 11 of the 15 admissions were in this group. They concluded that patients not transported were at risk of serious or fatal outcomes, and that although causation had not been established, EMS refusal to transport was associated with higher risk to the patient. They noted that their
TABLE3. Chief Complaints of Patients in Whom Follow-Up Was Obtained According to Whether They Sought Additional Medical Care Chief Complaint
Care Obtained (n = 95)
No Care Obtained (n = 104)
Altered mental status Cardiac/respiratory Diabetic-related Gastrointestinal Miscellaneous Musculoskeletal Neurological Obstetric/gynecological Psychiatric Intoxication (incl. ethanol) Trauma
1% 20% 0% 3% 8% 5% 10% 3% 1% 1% 48%
1% 10% 2% 5% 10% 7% 13% 0% 2% 3% 47%
NOTE: Numbers in each box expressed as percentage of total patients in that column.
TABLE4. Chief Complaint and Ultimate Hospital Discharge Diagnosis of Patients Who Initially Refused Medical Care But Later Were Admitted to Hospital Case ID
Chief Complaint
28 39
Cardiac/respiratory Trauma
54
Miscellaneous
93 113 147
Miscellaneous (weakness/nausea) Cardiac/respiratory Cardiac/respiratory
188 244 283 289 375 419 436
Cardiac/respiratory Obstetric/gynecological Cardiac/respiratory Cardiac/respiratory Trauma Trauma Diabetic-related
Discharge Diagnosis Myocardial infarction Diabetes mellitus and extremity trauma Gastroenteritis, dehydration Digoxin toxicity (level = 7.5) Suicidal ideation Congestive heart failure (died) Multiple sclerosis Normal vaginal delivery Myocardial ischemia Congestive heart failure Myocardial ischemia Rheumatoid arthritis Congestive heart failure
findings indicated the need for further, prospective study of the risks of nontransport. Sucov et al 2 retrospectively found 188 cases of patients who refused medical assistance in the out-of-hospital setting. No patients were refused transport by the EMS crews. By telephone follow-up, they were able to determine outcomes in 50% of these patients. Of these patients, 31 contacted a physician and 6 were admitted to a hospital. They note that this number probably underestimates the actual number admitted because of the relatively low follow-up achieved. They call for "additional studies to provide a better description of the natural outcome" of patients after RMA. Cone et aP 1 retrospectively analyzed 81 cases of patientinitiated refusal of prehospital care, obtaining a follow-up rate of 67% (54 patients). They report that 32% of these patients sought further care after RMA, with 7 (13%) admitted to the hospital. There were no deaths in their study group. They noted that on-line medical control appeared to improve documentation in cases of RMA, and called for research into the effects of physician contact on RMA. Alicandro et ali0 found that physician contact improves the transport rate of patients who attempt RMA. Our study is the first prospective study to address the issue of the natural outcome of patients after RMA. Since our system does not permit EMS refusal of transport, our study population is comparable to that of Sucov et al. 2 We found that 48% of the patients we followed up sought further care after RMA. This is comparable to the previously reported 32% to 64% from the retrospective studies. 1,2,~1 Thirteen patients in our study were admitted to the hospital; this 7% admission rate is somewhat lower than the other studies reported. It is possible that our study design was able to find more nonadmitted patients for follow-up, or that greater effort was made by out-of-hospital providers to transport ill patients in this prospective study. Although the number of patients is too low to be definitive, we note that 6 of 13 admitted patients had cardiac or respiratory chief corn-
26
AMERICAN JOURNAL OF EMERGENCY MEDICINE • Volume 14, Number 1 • January 1996
plaints, suggesting the possibility that these patients are at increased likelihood of requiring admission after RMA. Our study has several weaknesses. We report on 321 patients with a 20% improvement on follow-up (50% to 62%) compared with the study of Sucov et al,2 but we note that 38% of patients in our study were still lost to follow-up. This group included patients with serious chief complaints, such as altered mental status, cardiac or respiratory problems, and trauma. It is possible that significant outcomes were missed in this group lost to follow-up. Even if we make the unlikely assumption that all of the patients lost to follow-up did not seek further medical care, 30% of our total patient cohort that refused medical assistance would still have sought further care. Our study also has an observer bias, because the EMS crews knew we were analyzing RMA and may have striven to transport ill patients. This would have resulted in an underestimate of the frequency of adverse outcomes in patients who refused medical assistance. Our population is also specifically suburban/rural, and results may not generalize to urban populations. We also could not determine whether delay in obtaining care resulted in any increase in morbidity. Patients who interact with the EMS system may require transport to a hospital for admission, may receive interventions in the ED after transport to prevent admission, or may not require transport at all. It is possible that some patients may be better served by RMA followed by care given by an office-based physician. Our results, however, indicate that patients who refuse medical assistance may not always be able to make the appropriate distinctions of medical necessity. Detailed guidelines need to be developed to determine which patients may be safely allowed RMA as opposed to those in whom great effort should be made to ensure transport to a hospital. Until such criteria are developed, efforts should be made to convince all patients who attempt RMA to be transported to the hospital. Intervention by an on-line medical control physician has been shown to increase transport rates of patients who initially desire RMA,10 and would therefore be an option for EMS systems. Further study to address these issues is warranted.
The authors acknowledge the participation of the following volunteer corps and EMS personnel: Bayshore-Brightwaters Rescue Ambulance: Steve Engelberg, EMT-CC, Tess Melquenn, EMT-P, and Ted Suida, EMT-P; Brentwood Legion Ambulance: Juan Espinosa, EMTCC, Susan O'Malley, EMT-CC; Central Islip- Hauppauge Volunteer Ambulance: Dorothy Norman, EMT-CC, Joseph Ortiz, EMT-D, Joseph Tricario, EMT-D, and Meredith Whipple, EMT-D; Commack Volunteer Ambulance: Nancy Wichtendahl, RN, EMT-CC; State University of New York at Stony Brook: Robert Delagi, EMT-P.
REFERENCES 1. Zachariah BS, Bryan D, Pepe PE, Griffin M: Follow-up and outcome of patients who decline or are denied transport by EMS, Prehospital Disaster Medicine 1992;7:359-363 2. Sucov A, Verdile VP, Garettson D, Pads PM: The outcome of patients refusing prehospital transportation. Prehospital Disaster Medicine 1992;7:365-371 3. Holroyd B, Shalit M, Kallsen G, et al: Prehospital patients refusing care. Ann Emerg Med 1988;17:957-963 4. Stark G, Hedges JR, Neely K, Norton R: Patients who initially refuse prehospital evaluation and/or therapy. Am J Emerg Med 1990;8:509-511 5. American College of Emergency Physicians: Continuous Quality Improvement in EMS. Dallas, TX, ACEP, 1992, pp 298- 303 6. Mottley L: Refusal of medical assistance in the field. In Kuehl AE (ed): Prehospital Systems and Medical Oversight. St. Louis, MO, National Association of EMS Physicians and CV Mosby, 1994, pp 375-380 7. Selden BS, Schnitzer PG, Nolan FX, Veronesi JF: The "nopatient" run: 2698 patients evaluated but not transported by paramedics. Prehospital Disaster Medicine 1991 ;6:135-142 8. Selden BS, Schnitzer PG, Nolan FX: Medicolegal documentation of prehospital triage. Ann Emerg Med 1990;19:547- 551 9. Goldberg RJ, Zautcke JL, Konigsberg MD, et al: A review of prehospital care litigation in a large metropolitan EMS system. Ann Emerg Med 1990; 19:557-561 10. Alicandro JM, Hollander JE, Henry MC, et al: Impact of interventions for patients refusing EMS transport. Academic Emerg Med 1995;2:480-485 11. Cone DC, Kim DT, Davidson SJ: Patient-initiated refusals of prehospital care: Ambulance call report documentation, patient outcome, and on-line medical command. Prehospital Disaster Medicine 1995; 10:3-9 12. Soler JM, Montes MF, Egol AB, et al: The ten year malpractice experience of a large urban EMS system. Ann Emerg Med 1985;14: 982-985 13. Goldstein A: EMS and the Law: A legal handbook for EMS personnel. Bowie, MD, Robert J. Brady Co, 1983, pp 69-82 14. Morgan DL, Wainscott MP, Knowles HC: Emergency medical services liability litigation in the United States: 1987 to 1992. Prehospital Disaster Medicine 1994;9:214-221