Appendix B ADDITIONAL MATERIAL SUBMITTED BY MITCHELL GOULD, EXECUTIVE DIRECTOR, GOULD CONVALESARIUM, BALTIMORE, MD. ITEM 1. SUMMARY, BY PATIENT, OF ONSET OF ILLNESS, NOTIFICATION OF PHYSICIAN AND INITIATION OF TREATMENT BY PHYSICIAN ITEM 1. SUMMARY BY PATIENT, OF ONSET OF ILLNESS, NOTIFICATION OF PHYSICIAN AND INITIATION OF TREATMENT BY PHYSICIANContinued ITEM 2. PHYSICIANS TREATING PATIENTS WITH SALMONELLA SYMPTOMS AT THE CONVALESARIUM ITEM 3. CLINICAL ASPECTS, OUTBREAK OF SALMONELLA ENTERITIS, GOULD CONVALESARIUM, JULY 1970, BY WILLIAM B. GREENOUGH, CHIEF, INFECTIOUS DISEASES DIVISION, THE JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE BACKGROUND The nature of infection by salmonella is well outlined in the recent publication by The Baltimore City Health Department*. Death can result from several causes and the very young, very old or chronically ill are particularly at risk to fatal outcome. If the organism enters the blood stream, the blood vessels, heart and any other vital organs can be infected. Such infections are characterized by high fevers and local signs referable to the particular organ or organs involved. More commonly infections are limited to the gut and the main problem is fluid loss due to diarrhea. Fever, chills and abdominal cramps often accompany such gastroenteritis. The causes of death in the case of primarily intestinal infection are all related to decreased blood volume leading to lowered blood pressure and poor perfusion of vital organs such as the brain, heart and kidneys or deranged electrolyte concentrations secondary to losses of potassium and bicarbonate. In the current outbreak the disease was primarily an enteritis. Whether blood stream invasion occurred and contributed to some of the deaths cannot be judged without blood cultures. This data was not available to me and could be derived only from the charts of hospitalized patients. SOURCES OF INFORMATION 1) Visit to Gould Convalesarium: Talks with Mr. Gould and members of nursing staff. 2) Review of 55 charts of patients involved in the outbreak at the Gould Convalesarium. 3) Review of one hospital record. 4) Review of one autopsy report. 5) Center for Disease Control Morbidity and Mortality Weekly Report Volume 19, Number 32, 314 (August 15, 1970). NATURE OF THE POPULATION AND EPIDEMIC The average age of the patients who became ill was 78 years. All of them had serious underlying illness. The most common variety was vascular with evidence of damage to brain and heart. During the 10 days from July 26 until August 3, 1970, 104 of 145 patients (72%) and 19 of 66 employees (29%) were known to have developed diarrhea. Salmonella enteritis was cultured from the stools of 25 patients and 17 employees. Twenty-five patients died with diarrheal symptoms (case fatality ratio-25%). There were no deaths among employees. There was no difference in the case fatality rate between those patients treated in the nursing home as compared to those transferred to hospitals (Table 1). CLINICAL CHARACTERISTICS OF ILLNESS Review of 55 charts of the most severely affected patients has indicated that although there were five deaths within 72 hours of the onset of diarrhea, the average time of death was 8.0 days after onset of diarrhea. Most of the patients dying more than one week after onset of illness had little or no diarrhea at the time of death. The average time after onset of diarrhea when patients were transferred to hospitals was 6.6 days and as noted above the mortality rate of the hospitalized patients was 6/28 or 22% which was identical with the 19/79 or 24% mortality seen in patients remaining in the nursing home. Of those patients with data available for blood pressure during the 24 hours before hospitalization only 2 out of 10 patients had sustained decreases of more than 20 mm of mercury over their normal systolic blood pressures. The population whose charts were reviewed in detail are described in Table 2. The characteristics of the illness are summarized in Table 3. All patients had diarrhea and the majority also had fever with the diarrhea. A decrease * Baltimore Health News, vol. XLVII, 121-124, Aug.-Sept. 1970. in systolic blood pressure of more than 20 mm of mercury occurred in 17 out of 32 patients at some time during their illness where there were sufficient recordings to determine the incidence of hypotension. Hypothermia of less than 97.5° F rectal temperature occurred in 5 out of 44 instances. TREATMENT In general, patients were treated by increasing their intake of oral fluids. In 2 out of 53 instances intravenous replacement with electrolyte solutions was carried out. In 2 further patients clysis with isotonic saline was given (Table 4). A large number of patients (28) were sent to hospitals when their condition seemed to indicate treatment beyond the capacity of the nursing home. In most cases referral to the hospitals was accomplished before a serious drop in blood pressure had occurred, 2 out of 10 where data was available. Most patients had very severe underlying vascular disease which had resulted in obvious brain or heart damage. Such patients were highly susceptible to any decrease in blood flow to vital organs. It is clear from the autopsy report of Mr. Robert Simpson, Johns Hopkins autopsy No. 37433 that dehydration, hypotension and resulting diminished perfusion of vital organs resulted ultimately in his death. DISCUSSION An overwhelming epidemic of diarrhea due to salmonella enteritis resulted in 25 deaths within a short period of time in The Gould Convalesarium. Although in the 5 cases who died within 72 hours and an undetermined number of other cases may have died because of loss of body fluids, it is striking that the majority of cases which died survived an average of 8 days. Thus the bulk of deaths cannot be attributed to simple dehydration and shock because of fluid loss. Furthermore, it is of interest that the mortality rate in those patients who were hospitalized was not significantly different from those remaining in the nursing home. This again suggests that complications related to the underlying diseases were more at fault then was simple dehydration. In otherwise healthy people with severe fluid loss due to diarrhea, replacement of the deficit, intravenously or by mouth, if shock is not present prevents all deaths. Hence, if fluid replacement was the main problem none of the hospitalized patients should have succumbed. This of course assumes that the hospitals gave the needed intravenous fluids which was true in the charts of the two hospitalized patients I have reviewed. On the other hand, if all the patients who experienced voluminous diarrhea had an accurate appraisal of their degree of dehydration by measurement of plasma protein or specific gravity and had prompt intravenous replacement heen given, it is possible that some of the complications of renal failure and cardiovascular failure might have been averted. It would not be possible to carry out such therapy in the setting of a nursing home unless a special team of physicians had been called in at the earliest indication of epidemic diarrhea. It is not likely that even with more prompt hospitalization of all cases that there would have been very much gained. The damage done by hypotension in such a population of patients is irreversible at a very early stage. Furthermore, in patients, many of whom already had neart failure, the rapid administration of intravenous electrolyte solutions is fraught with the great risks of pulmonary adema and death if used injudiciously and without proper control. RECOMMENDATIONS FOR THE FUTURE I believe the only really effective way to reduce mortality in an explosive outbreak of diarrheal illness such as that which occurred at The Gould Convalesarium is to know ahead of time who to call on for emergency assistance. In Baltimore there are two groups of physicians expert in the management of severe dehydrating diarrhea who could be called on to render the needed measurements and guide treatment as soon as an outbreak occurs. These are the group of The Infectious Diseases Division of The University of Maryland Medical School under Dr. Richard Hornick and the group of the Infectious Diseases Division of The Johns Hopkins University School of Medicine which is under my direction. Short of getting prompt help I think there could be little improvement over the performance of the staff of the Gould Convalesarium who actually managed a very large epidemic in an admirable fashion in light of the limited staff available. SUMMARY (1) An explosive epidemic of dehydrating diarrhea did occur and can be documented from the records. (2) When blood pressures were found to be low, patients were transferred to hospitals. (3) Many patients were critically ill before the onset of diarrhea and although fluid loss hastened death it cannot be incriminated as the sole cause of death. (4) People without vascular disease survived this strain of salmonella enteritis readily. The mortality mainly reflects the severity of the vascular disease in the older patients. (5) Earlier recognition of the extent of fluid losses and more prompt hospitalization and intravenous therapy would have been helpful but would not have saved all the patients involved because of the reasons cited in 4. (6) Laboratory measurements by which early detection and proper control of intravenous fluid therapy are the province of a hospital and should not be expected in nursing homes. (7) In future outbreaks of this severity prompt assistance with measurements of dehydration and guidance in replacement therapy would be the most effective means to reduce mortality and morbidity. Both the University of Maryland and The Johns Hopkins University have groups of physicians possessing a large experience and high degree of current skill in managing diarrheal illness. These resources may be called in future outbreaks. TABLE 1. CHARACTERISTICS OF PATIENTS HOSPITALIZED VERSUS THOSE NOT HOSPITALIZED Hospitalized from this group, outcome of these not known, 2 charts Fever 100 degree F. (no data 10). Drop in blood pressure >20 mm. Hg. over lowest value_ TABLE 4.-Treatment Intravenous Ringer's lactate or saline.... Clysis saline---- 2/53 2/53 |