BRBEF REPORT

Salmonella poona infection and surveillance in a neonatal nursery Anne Stone, RN, BSa Maria Shaffer, FIN’ Robert L. Sautter, PhDb Harrisburg, Pennsylvania

Background: This article reports the first known outbreak of Salmonella poona infection in a neonatal unit. Three babies had stool cultures positive for the organism. At the same time, S. poona was the cause of a nationwide food-borne outbreak associated with contaminated canteloupe. To minimize the neonatal outbreak, surveys were performed and control measures were instituted. Methods: Epidemiologic surveillance included the culture of rectal swabs collected from 48 employees, 18 babies, and four family members of the babies. Control measures used included the placement in cohorts and isolation of infected babies, strict adherence to universal precautions, and the restriction of visitation in the nursery. Results: A total of three babies and one employee in the surveillance were found to have Salmonella sp. An additional two hospitalized adult patients had S. poona. Of all the people included in the surveillance, only the three babies were found to have S. poona. The hospital employee had S. enteritidis. Conclusions: Timely culture results, rapid cohort placement of infected babies, and strict adherence to universal precautions led to the successful eradication of the organism. (AJIC AM J INFECT CONTROL 1993;21:270-3)

Hospital outbreaks of salmonella are common. 1 Frequent vehicles of transmission have been food or food handlers,2-6 inanimate environment,7710 and person-to-person contact. l-3. lo Salmonella in- fections are uncommon in neonatal nurseries: however, several infectious outbreaks have been documented.7~8~ ‘l-l5 The serotypes reported are Salmonella typhimurium,‘3 Salmonella enteriti- dis,7* ” Salmonella anatum,12 Salmonella cuba- na,14 Salmonella weltervreden,‘4 Salmonella new- PO?-& I2 Salmonella bareilly,12 Salmonella oranien- berg,12 Salmonella senftenberg,14 Salmonella worthington,7 Salmonella eimsbuettel,’ and Sal- monella heidelberg.” Documented modes of trans- mission have been rectal thermometers8 in- adequate handwashing,‘, lo* l6 oropharyngeal suc- tion,’ infected mothers, “* l8 and contaminated

From the Department of Nursing, Division of Infection Control,a and the Departments of Pathology and Microbiology,b Harrisburg Hospital, Harrisburg, Pennsylvania.

Reprint requests: Robert L. Sautter, PhD, Department of Microbi- ology, Harrisburg Hospital, S. Front St., Harrisburg, PA 17101-2099. 0 1993 by the Association for Practitioners in Infection Control, Inc.

0196-6553193 $01 .OO + 0.10 17/47/47809

270

equipment. lo Many outbreaks associated with the nursery, as well as other wards of the hospital, have failed to document a source of the infec- tion.‘O In addition, pseudoepidemics have been re- ported involving a contaminated rubber pipette bulb in the laboratory19 and failure to routinely serotype isolates in the clinical laboratory. This incident involves infection with Salmonella poona (group G) and subsequent surveillance in a neona- tal intensive care unit (NICU).

Symptoms of salmonella infection in neonates have a varied range of presentation, from no symptoms to fever, diarrhea, jaundice, septicemia, and death.‘, ‘9 * All patients who have received antibiotics, antacids, and immunosuppressive drugs are at risk for severe symptoms.’ Premature neonates, who are highly immunocompromised individuals, are at increased risk for serious disease associated with salmonella infections, with consequences that may include death. ‘, ’

During the summer of 1991, a nationwide outbreak of S. poona infections was documented as associated with contaminated canteloupe. Twenty-six cases were reported to the Pennsylva- nia Department of Public Health as part of the

AJIC Volume 21, Number 5 Stone, Shaffer, and Sautter 271

investigation. To our knowledge, this is the first reported outbreak of S. poona infection in a neonatal unit.

MnNoDS Neonatal Unit

The unit involved is located in a special wing of the women and children’s department of a 465- bed general hospital. All neonates weighing less than 2000 gm or of any weight and requiring intensive neonatal care are admitted to this unit. Visitors are restricted to parents, grandparents, and siblings older than 12 years. Siblings younger than 12 years need special permission from the neonatologist to enter the unit. There is a single main entrance to this unit, with an anteroom that provides gloves, cover gowns, masks, and hand- washing facilities. The unit has a 22-bed capacity, with incubators and specialty units available for each neonate.

St8tf This unit is staffed by a 45-member specialty

neonatal nursing staff and three on-site neonatol- ogists. Typically, one nurse is assigned to care for two or three babies. Staffing is adjusted for neonates that require one-on-one care.

RESULTS

Surveillance efforts were complicated by the presence of S. poona infection in two adult patients. Patient 1, admitted June 4, 199 1, proved to have a community-acquired infection. The second adult (patient 2) was a 39-year-old insulin- dependent diabetic in premature labor at 26 weeks. This patient was first admitted to the hospital on June 18, 199 1. The two adult patients were physically separated on two different units; patient 1 was on a medical-surgical unit and patient 2 was on an alternative maternity floor. The only common staff between these two units were dietary porters. Careful review of the health history of the dietary personnel assigned to these areas revealed no apparent vehicle for infection. Stool cultures collected from these employees and from two cooks were negative for enteric patho- gens. Chart review of patient 2’s delivery revealed minimal episodes of diarrhea. During this time, she received only intravenous fluids and medica- tions. On June 24, this patient had precipitous and spontaneous delivery of a premature female child. The baby was delivered with intact membranes and transferred to the NICU. At that time the mother (patient 2) started to have multiple epi-

Tahbk 1. Surveillance data and associated control measures

control lll666W66 R66Ult6

Second neonatal case detected

Isolation and cohort placement of two babies within the NICU

Formation of a multidisciplinary task force to address problem

Surveillance of 18 babies and 48 staff members

An additional baby was positive for S. poona

All health care personnel were nega- tive for S. poona

One nurse was off sick with a gas- trointestinal illness caused by S. enteriditis

The breast milk from mothers of in- fected babies and baby formula were culture negative for salmo- nella

By day two of the The three infected neonates were outbreak, three transferred from the NICU to a sat- babies were found ellite location on a separate floor positive for S. poona

Additional measures Any new admissions to the NICU were placed in a separate area in the unit

All surfaces and equipment were cleaned and disinfected

Visitors in the main NICU were re- stricted to parents only

sodes of diarrheal stools. On June 25, patient 2 became febrile, with a maximum temperature of 39.4” C. A stool specimen was obtained on June 26, with results available on June 28 positive for the presence of S. poona.

During the fourth week in June 1991, the microbiology laboratory had isolated “colonies suspicious of salmonella” from a culture of stool collected from an infant in the neonatal unit. S. poona was identified in this culture from the apparently septic 8-week-old baby (index baby). Blood cultures were also positive for S. poona. This index baby was not related to either adult patient with S. poona. The infant was isolated within the NICU and started on a regimen of ampicillin, gentamicin, and cefoxitin (Table 1).

Adult patient 2, who was discharged, was subsequently readmitted to the hospital with se- vere diarrhea and S. poona sepsis. Her infant, who was still in the NICU, had no symptoms and was

272 Stone, Shaffer, and Sautter AJIC

October 1993

progressing well and gaining weight. This infant was isolated and observed for signs and symp- toms of infection. Stool cultures were positive for S. poona but the infant remained free of symptoms.

A rectal swab was collected from each baby in the nursery and sent to the microbiology labora- tory. One additional baby was found to have S. poona. This infant had already been isolated because of suspicious, foul-smelling stools, fever, and sudden onset of apnea and bradycardia.

Surveillance of the staff in the neonatal unit found one nurse who had been ill after assignment to the index baby. She had been off duty since the first day of the outbreak and had positive stool cultures from a different laboratory for SaZmo- nella sp. Initial evidence of a diarrhea-like illness made her a likely candidate as the potential point source of the outbreak. Subsequent evaluation by the reference laboratory and the hospital labora- tory identified the organism as Salmonella enter- itidis. According to employee health guidelines at the hospital, the nurse was required to have three consecutive negative stool cultures before return- ing to work.

COMMENTS

Determination of the point source of our inci- dent was complicated by the following facts: (1) a nurse in the neonatal unit as well as a postpartum mother of an apparently well baby both exhibited gastrointestinal symptoms and (2) during the summer of 1991, an outbreak of S. poona was identified, ranging geographically to include Texas, New York, and Pennsylvania. Public health alerts connected this statewide outbreak with contaminated cantaloupe melon. Careful exami- nation and follow-up of the infected adults and employees in our outbreak implicated the infected mother as the most likely point source for the infection. The ill nurse had been infected with a different serotype of salmonella.

Silent maternal transmission of salmonellosis has been previously documented.14* “8 l8 In these outbreaks, contributing factors in the prenatal period of the index mother included premature rupture of the membranes and premature deliv- ery. Our adult index case had premature labor and delivered a premature female infant. Even though her membranes were intact, studies have shown that a significant number of infected mothers can pass salmonella to their neonates, as apparently happened in this situation.18 The index adult case

was an insulin-dependent diabetic woman who had symptoms during her first admission in the hospital. Her symptoms worsened during a pro- longed period and necessitated a second admis- sion to the hospital. No point source of infection while she was in the hospital could be identified. The most likely conclusion is that this patient had been infected with a low dose of the organism before admission. Diabetes may have been a predisposing factor that lowered the dose of inoculum required to cause infection and possibly prolonged the demonstration of symptoms and complicated the detection of the source of the incident.lS

Once introduced in the nursery, S. poona was transmitted from infant to infant by way of personnel. This may have occurred during emer- gency cardiac stimulation, when personnel did not wash hands between patients.

Prompt reporting and identification of isolates in the laboratory and immediate action by the ICPs in developing a multidisciplinary task force to handle the problem proved to be the control measures that prevented an outbreak. In other studies,8* l5 even a slight delay in recognition or action resulted in more infected babies. It is difficult to compare outbreaks involving different serotypes of salmonella, however, because some of the strains have a greater propensity to cause infection than others.’

The creation of a separate satellite NICU, which isolated the infected babies, prevented closure of the main NICU and contained the spread of infection of S. poona. The isolation of neonates in a unit separate from the main nursery has previ- ously proved an effective infection control prac- tice.*l Investigation, universal control measures, placement in cohorts of the infected infants, and the use of strict handwashing resulted in contain- ing the spread of a potentially serious infection in the NICU.

Infection control recommendations after the outbreak were as follows: (1) strict universal precautions for all neonates, (2) enforced hand- washing and gloving in the NICU, (3) restriction of staff and visitors with existing respiratory or gastrointestinal illness, (4) maintenance of an “early warning” alert system between the labora- tory and the ICPs, and (5) open communications to staff and parents of the babies. By adhering to these recommendations, it is our contention that the NICU will be able to prevent outbreaks of this nature in the future.

AJIC Volume 21, Number 5

We thank the microbiology laboratory and the NICU staff for their role in identifying the isolates and caring for the babies. We also thank the multidisciplinary task force and nursing administration for their contribution to the timely control of this situation. Finally, we thank Drs. Ernest J. Davis, Richard D. Baltz, and Richard N. Blutstein for their technical assis- tance.

References

5.

6.

7.

8.

9.

10.

Weikel CS, Guerrant RL. Nosocomial salmonellosis. Infect Control 1985;6:218-20. Cruickshank JG. The investigation of salmonella outbreaks in hospitals. J Hosp Infect 1984;5:241-3. Palmer SR, Rowe B. Investigation of outbreaks of salmo- nella in hospitals. BMJ 1983;287:891-3. Pollock AM, Whitty PM. Crisis in our hospital kitchens: ancillary staffing levels during an outbreak of food poison- ing in long stay hospital. BMJ 1990;300:383-5. Spitalny KC, Okowitz EN, Vogt RL. Salmonellosis out- break at a Vermont hospital. South Med J 1984;77: 168-72. Wilkinson PJ. Food hygiene in hospitals. J Hosp Infect 1988;11:77-81. Khan MA, Abdur-Rab M, Israr N, et al. Transmission of Salmonella worthington by oropharyngeal suction in hos- pital neonatal unit. Pediatr Infect Dis J 1991; 10:668-72. McAllister TA, Marshall A, Roud JA, Holland BM, Turner TL. Outbreak of Salmonella eimsbuettel in newborn infants spread by rectal thermometers. Lancet 1986; 1: 1262-4. Dwyer DM, Klein EG, Istre GR, Robinson MG, Neumann DA, McCoy GA. Salmonella newport infections transmitted by fiberoptic colonoscopy. Gastrointest Endosc 1987;33: 84-7. Joseph CA, Palmer SR. Outbreaks of salmonella infection in hospitals in England and Wales 1978-87. BMJ 1989; 298:1161-3.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Stone, Shaffer, and Sautter 273

Bannerman CH. Salmonella heidelberg enteritis an out- break in the neonatal unit Harare central hospital. Centr Afr J Med 1985;31:1-4. Aggarwal P, Sarkar R, Singh M, Anand BRG, Chowdhuri ANR. Salmonella bareilly infection in a pediatric hospital of New Delhi. Ind J Med Res 1983;78:22-5. Chaturvedi P, Narang P, Khan MJ, Sharma S. An outbreak of Salmonella typhimurium infection in a paediatric ward. Ind J Path Microbial 1985;28:121-8. Joseph AT, Rammurty DV, Srivastava L, Gupta R, Mohan M, Anand NK. Salmonella senftenberg outbreak in a neonatal unit. Ind Pediatr 1990;27:157-60. Seals JE, Parrott PL, McGowan JE, Feldman RA. Nursery Salmonellosis: delayed recognition due to unusually long incubation period. Infect Control 1983;4:205-8. Tauxe RV, Hassan LF, Findeisen KO, Sharrar RG, Blake PA. Salmonellosis in nurses: lack of transmission to patients. J Infect Dis 1988; 157:370-3. Baxter DN, Morton S. An outbreak of S. indiana in a maternity unit: implications for control policies. Commun Med 1987;9:365-71. Roberts C, Wilkins EGL. Salmonella screening of pregnant women. J Hosp Infect 1987;10:67-72. Harris AA, Pottage JC, Fliegelman R, et al. A pseudoepi- demic due to Salmonella typhimurium. Diagn Microbial Infect Dis 1983; 1:335-7. Baddour LM, Robinson VL, Baselski V. Pseudoepidemic of salmonellosis in a nursery: importance of isolate serotyp- ing. AM J INFECT CONTROL 1987;15:79-80. Lamb VA, Mayhall CG, Spadora AC, Markowitz SM, Farmer JJ III, Dalton HP. Outbreak of Salmonella typh- imurium gastroenteritis due to an imported strain resis- tant to ampicilin, chloramphenicol, and trimethoprim- sulfamethoxazole in a nursery. J Clin Microbial 1984;20: 1076-9.