Nosocomial Septicemia and Meningitis in Neonates

CAROL 1. BAKER, M.D.*

tlouston. Texas

From the Departments of Pediatrics, Microbiol- ogy and Immunology, Baylor College of Medi- cine, and lefferson Davis Hospital, Houston. Texas. This study was presented at the 2nd In- ternational Conference on Nosocomial Infec- tions, held August 5-8.1980. in Atlanta, Georgia. Requests for reprints should be addressed to Dr. Carol J. Baker, Department of Pediatrics, Baylor College of Medicine, 1200 Moursund Avenue, Houston, TX 77030.

* Recipient of U.S.P.H.S. Research Career Development Award 1 K04 AI 00323 from the National Institute of Allergy and Infectious Diseases.

The past decade brought with it both highly sophisticated neonatal intensive care with improved perinatal mortality rates, and increased risk for nosoco.mial septicemia and meningitis among survivors. Although most of these infections were caused by multiple antibi- otic-resistant gram negative enteric bacteria, Staphylococcus aureus “outbreaks” appeared sporadically. Risk for nosocomial infection was related primarily to factors which enhance infant contact with these bacteria (crowding and high nurse to infant ratios, poor handwashing practices, contaminated life support equipment, an- tibiotic exposure and prolonged hospitalization) in combination with those poorly defined determinants of bacterial virulence and host defense. Control measures for the prevention or reduction of these infections are aimed at decreasing the neonate’s contact with the “outbreak” strains-improvement in handwashing practices and equipment sterilization processes, cohorting of infants, establishment of appropriate nurse:infant ratios, and, in certain instances, modi- fication of antibiotic practices. Future efforts should be directed toward better definition of bacterial virulence, host susceptibility and preventive measures.

The first 28 days of life (the neonatal period) is a time of unprecedented risk for the development of bacterial septicemia and meningitis. This fact has not been influenced to any degree by the advances recorded during the past decade by the disciplines of obstetrics, neonatology, microbiology or antimicrobial pharmacology, but the substance of this fact has changed. Although the majority of bacterial infections among neonates still occur during the first few days of life and are the result of intrapartum exposure to maternal genital microorganisms, an in- creasing percentage of these infections have their onset beyond the age usually required for newborn hospitalization (the first three days of life] and are due to nosocomial pathogens. This latter fact is a result of the tremendous advances in neonatal intensive care which have resulted in both a decrease in perinatal mortality, especially among low birth weight infants, and an increased risk for septicemia and meningitis caused by microorganisms resident in the nursery among survivors. The past 10 years has given to those involved with hospital infection control programs objective evidence for the significant contribution which nosocomial septicemia and meningitis among neonates makes to infant mortality and morbidity in this country and a challenge to define basic host-bacterial interactions which allow their occurrence. Only through an understanding of pathogenesis will it be possible to realize the goal of effective methods for the reduction and, ultimately, the prevention of these infections in the future.

EI’IOLOGIC AGEWIS

The neonate has certain characteristics with regard to nosocomial septicemia and meningitis, but there are many more parallels with the

008 March 1981 The American Journal of Medicine Volume 70

older patient when considering etiologic agents and risk factors which underlie these infections in neonates. In Table I are summarized the microorganisms which have been implicated in hospital-acquired “outbreaks” of neonatal septicemia and/or meningitis since 1970. Whether this list is comprehensive and accurately proportioned or whether it merely reflects methods of reporting cannot be known. However, it is clear that almost any microorganism, given the appropriate cir- cumstances, has the capacity to produce disseminated infection in the neonate and that during the past 10 years, the majority of nosocomial infections in neonates have been due to gram-negative enteric bacteria (ap- proximately 70 to 80 percent) [l-3]. When methods for the serologic classification of these agents have been available, virtually all of the “outbreaks” have been associated with a single serotype [5,7,9,12,15]. Whether this observation is related to increased virulence of certain strains for the neonate or to increased preva- lence in the nursery environment, or both, cannot be accurately assessed at this time. In one study, increased lethality in mice was produced by the “epidemic” strain of type 33 Klebsiella pneumoniae when compared to other Klebsiella isolates from neonates [9]. “Virulence” of certain strains may also be related to their ability to develop and/or transfer resistance genes which is well illustrated by the usual occurrence of multiple antibi- otic-resistant gram-negative enteric bacilli as predom- inant nosocomial pathogens in neonatal intensive-care settings [1,3,10]. Recently, new patterns of resistance to methicillin or aminoglycosides in Staph. aureus isolated from neonates have been recognized [1,5]. The “viru- lence” factors, which account for the ability of these reported nosocomial pathogens to gain access to the bloodstream of the neonate, are not defined. However, in contrast to the older host, this bacteremia in the neonate commonly results in meningeal invasion (ap- proximately 25 percent) with its concomitant increase in mortality and morbidity.

The diversity of etiologic agents and their antimi- crobial susceptibility patterns which are recorded for neonates in individual hospitals is not found when an- alyzing the events which lead to an “outbreak” of sep- ticemia and/or meningitis. These events involve the introduction of the microorganism into the nursery, its transmission to multiple infants, its ability to adhere to and multiply on (colonize] multiple mucous membrane sites, and its penetration into the bloodstream with the resultant production of symptomatic disease. The source from which a nosocomial pathogen is introduced into the nursery is usually not known. For gram-negative enteric bacilli, introduction commonly occurs from an individual infant “contaminated” by his antibiotic-al- tered bowel flora, an outside hospital or improperly sterilized life-support equipment. Introduction by nursery personnel, common in infections due to Staph. aureus and group A Streptococcus [4,6,21] is rare for gram-negative enterics. The evaluation of many “out-

TABLE I Etiologic Agents Associated with Nosocomial SepticemialMenlngitis in Neonates

Microorganism

Gram-positive Staphylococcus aureus Group A Streptococcus Groups B and D streptococci

Gram-negatives Klebsiella pneurponiae Proteus mirabilis Enterobacter cloacae Serratia marcescens Citrobacter diversus Citrobacter koseria Pseudomonas aeruginosa

References

[=I [Sl [3,7,81

La;“’ [151

[I'31

;:z:j

breaks” and the data from considerable clinical research in the past 10 years makes it clear that the major reser- voir for nosocomial microorganisms in the nursery en- vironment is the colonized neonates themselves. Once introduction has occurred, however, transmission of the agent from colonized neonates via the hands of nursery personnel to noncolonized infants is the primary vehicle by which dissemination and propagation of a particular strain occurs in a given nursery. This observation applies to both gram-positive and gram-negative bacteria. Respiratory transmission of bacteria causing neonatal septicemia and meningitis has not been documented [1,211.

Since the neonate at risk for serious nosocomial in- fection has a limited ability to respond to bacteremia with the usual clinical signs, suspected infection re- quires prompt initiation of empirically selected anti- microbial agents. Their appropriate selection will de- pend upon a knowledge of the prevalent nosocomial pathogen(s) in a particular nursery. In our nursery, as in most, a single gram-negative enteric organism (Klebsiella pneumoniae) which is resistant to the ami- noglycoside “routinely” employed for the treatment of suspected early onset (age less than three days], mater- nally-acquired sepsis will predominate at a given time. Knowledge regarding prevalent nosocomial pathogens and their antimicrobial susceptibility can only result from continuous monitoring of microbiologic data by infection control personnel and dissemination of this information to professionals responsible for neonatal care. Those documented risk factors associated with nosocomia1 septicemia and meningitis among neonates should also be known so that a high index of suspicion will be apparent for the infant with nonsp&ific clinical signs and multiple risk factors. These include prema- turity with its concomitant immature immunologic de- fense mechanisms and associated illness (hyaline membrane disease, congenital malformations, birth asphyxia) [3], prior exposure to antimicrobial agents [9,10,13,15], prolonged hospitalization [3,15], contami-

NOSOCOMIAL SEPTICEMIA AND MENINGITIS IN NEONATES-BAKER

March 1981 The American Journal of Medicine Volume 70 899

NOSOCOMIAL SEPTICEMIA AND MENINGITIS IN NEONATES--BAKER

TABLE II Nosocomlal Septlcemla and Meningltls at Jefferson Davis Hospltal, 1975-1979

Yecv Total Caws’

@.I Total Live Rklhs

(No.)

Attaclc Rate: Total CaeesI 1,~

Llve Birthr hlecllom Ac@rad allor Bklh No. %

1975 70 8.830 8.1 19 27.1 1976 77 9,481 8.1 23 29.9 1977 68 10,442 6.5 22 32.4 1978t 71 11,911 6.0 16 22.5 1979 52 13,153 4.0 5 9.6

l Includes cases of infection acquired from maternal as well as nursery exposures. t In May. infection control policy program completed.

nated life support equipment such as intravenous catheters, transducer devices, respiratory equipment and intravenous solutions, as well as conditions which enhance infant contact with multiple antibiotic-resistant gram-negative enteric bacteria (crowding, high nurse to infant ratios, poor handwashing). It is apparent that several of these factors are interrelated, and the precise contribution of a single determinant of risk in an indi- vidual patient is usually impossible to assess.

METHODS FOR CONTROL AND PREVENTION

Control and prevention of nosocomial disease in neo- nates should be the primary goal of hospitals housing neonatal intensive-care units. Surveillance data re- garding disease rates per month as well as by site and pathogen should be reviewed by the hospital’s infection control committee. Such committees should be com- posed of a teain of persons capable of interpreting these data (physicians trained in infectious disease and neonatology, a microbiologist, the infection control nurse) as well as of making and implementing recom- mendations (nursing personnel, administrators and, upon occasion, representatives of speciality departments such as housekeeping, laundry and the like.)

The mainstay of prevention of nosocomial infection in the newborn nursery is provision of conditions which minimize manual contact in the transmission of bacteria [2l]. Adequate and convenient handwashing facilities must be available, and a rigidly enforced policy for personnel to wash with one of several antiseptic skin cleansers immediately before handling an infant is mandatory. Periodic in-service education programs serve to reinforce such policies. Inadequate nursery space and especially inadequate personnel can lead to breaks in technique. The American Academy of Pedi- atrics recommends one nurse for every one or two in- fants in intensive care, one for every three or four for intermediate care, and one for every six to eight for normal newborn nurseries [22].

To reduce the risk of contamination, optimal proce- dures for disinfecting life support equipment should be identified, publicized, enforced, and monitored by pe- riodic surveillance. Everything that comes in contact with the neonate must be discarded or disinfected be- fore use by another. In addition, nurseries should be

cleaned with equipment not employed in other parts of the hospital.

When an increased incidence of infection is detected, information should be quickly gathered to ascertain whether there is a common etiologic agent and possible clues regarding its source and transmission. Infants with known or suspected infection should be isolated, and their contacts within the nursery should be cohorted until the time of hospital discharge. Aseptic techniques, especially handwashing practices, should be reviewed and corrected if deficient. In “outbreaks” due to enteric bacilli, cultures of specimens from equipment and en- vironmental sources may be helpful upon occasion, and periodic cultures of swabs from throat and rectal sites in neonates for isolation of the “epidemic” agent may assist in cohorting of asymptomatically colonized in- fants.

Since June 1978, the newborn nurseries at Jefferson Davis Hospital in Houston, Texas, have followed a policy of routine cohorting of infants. This policy has been accompanied by an active infection control com- mittee which meets monthly to review surveillance data. routine in-service education of nursing and physician personnel regarding the importance of manual trans- mission of bacteria in the occurrence of nosocomial infections, and the voluntary control of antimicrobial usage in neonates through physician education pro- grams. The results of such a combined approach in a hospital in which high risk parturients, who are often indigent, are cared for is illustrated by the marked re- duction in nosocomial septicemia and meningitis re- corded for the year, 1979 (Table II]. To put these results in proper perspective, 47 infants had septicemia and/or meningitis due to maternally-acquired organisms (17

group B streptococci, 12 Esch. coli, 14 streptococci not group B, two gram-negative enterics. one Listeria mo- nocytogenes, one Hemophilus influenzae, nontypable) among the 13,153 births during 1979. but only five were attributed to nosocomial pathogens.

PROSPECTS FOR THE FUTURE

Although the past decade has allowed the accumulation of a large body of knowledge concerning nosocomial septicemia and meningitis in neonates, much of our current information regarding predominant etiologic

700 March 1981 The American Journal of Medicine Volume 70

NOSOCOMIAL SEPTICEMIA AND MENINGITIS IN NEONATES-BAKER

agents, epidemiologic determinants of risk, modes of In particular, identification of effective education pro- transmission and methods useful for control of “out- grams for nursery personnel emphasizing the impor- breaks” is purely descriptive. Future efforts should at- tance of handwashing practices and cohorting systems; tempt to delineate mechanisms by which certain for hospital administrators, stressing the absolute re- microorganisms acquire properties enhancing virulence quirement for adequate space and personnel in neo- for some neonates and by which the majority of infants natal special care units as well as financial support for exposed to these agents successfully evade their pene- infeciion control activities; and for physicians, indicating tration into the bloodstream and meninges. The latter the relative risk attributable to commonly employed should define the role of neonatal defense mechanisms invasive procedures in the development of bacteremia such as specific antibody, classic and alternative com- due to nosocomial pathogens and the proper use of plement pathways, leukocyte function and cell-med- empiric anticrobial agents for suspected infection. The iated immunity as they pertain to nosocomial bacteria. acceptance of infection control programs achieved In addition to basic investigation regarding the patho- during the past 10 years must be continued and ex- genesis of nosocomial infection in the neonate, intense panded. Only in this manner will progress continue. In effort should be given to the performance of carefully addition, hospitals themselves-not just university af- designed studies to evaluate the efficacy and cost-ef- filiated investigators and governmental health agen- fectiveness of intervention methods likely to prevent a ties-must bear some of the “burden” necessary to fu- major portion of nosocomial disease among neonates. ture investigation.

REFERENCES

5.

6.

a.

9.

10.

11.

12.

Gardner P, Bennett JV, Burke JP. McGowan JE, Jr, Wenzel RP: Nosocomial management of resistant gram-negative bacilli. J Infect Dis 1980; 141: 415-417.

Finland M: Emergence of antibiotic resistance in hospitals. Rev Infect Dis 1979; 1: 4-21.

Hemming VG, Overall JC, Jr. Britt MR: Nosocomial infections in a newborn intensive-care unit. N Engl J Med 1976; 294: 1310-1316.

Gezon HM, Thompson DJ, Rogers KD, Hatch TF, Rychek RR, Yee RB: Control of staphylococcal infections and disease in the newborn throueh use of hexachloronhene bathing. Pediatrics 1973; 51: 3%-339.

Faden H, Neter E, McLaughlin S, Giacoia G: Gentamicin- resistant Staphylococcus aureus: emergence in an intensive care nursery. JAMA 1979; 241: 143-145.

Geil CC, Castlk $VK, Mortimer, EA, Jr.: Group A streptococcal infections in newborn nurseries. Pediatrics 1970; 46: 849-854.

Steere AC, Aber RC, Warford LR, et al.: Possible nosocomial transmission of group B streptococci in a newborn nursery. J Pediatr 1975; 67: 784-787.

Aber RC, Allen N. Howell JT, Wilkinson HW, Facklam RR: Nosocomial transmission of erouo B strentococci. Pediatrics

Y . 1

1976; 56: 346-353. Hable KA, Matsen JM, Wheeler DJ, Hunt CE, Quie PG:

Klebsiella type 33 septicemia in an infant intensive care unit. I Pediatr 1972; 80: 929-924.

Eisenbach KD, Reber RM, Eitzman DV. Baer H: Nosocomial infections due to kanamycin-resistant R-factor carrying enteric oreanisms in an intensive care nursers. Pediatrics 1972; 50: 3u95-402.

,

Kayyali MZ, Nicholson DP, Smith IM: A Klebsiella outbreak in a pediatric nursery: emergency action and preventive surveillance. Chin Ped 1972; 11: 422-426.

Hill HR, Hunt CE, Matsen JM: Nosocomial colonization with

13.

14.

15.

16.

17.

la.

19.

20.

21.

22.

Klebsjella, type 26, in a neonatal intensive-care unit as- sociated with an outbreak of sepsis, meningitis, and nec- rotizing enterocolitis. J Pediatr 1974; 65: 415-419.

Montgomerie JZ: Epidemiology of Klebsiella and hospital- associated infections. Rev Infect Dis 1979; 1: 736-752.

Burke JP, Ingall D. Klein JO, Gezon HM, Finland M: Proteus mirabilis infections in a hospital nursery traced to a human carrier. N Engl J Med 1971; 264: 115-121.

Baker Cl. la&son CV. Kennv ID: An outbreak of seDticemia and n&ngitis due’to Entkiobacter cloacae in a neonatal intensive-care unit (abstract). Proceedings of the lath In- terscience Conference on Antimicrobial Agents & Che- motherapy, held in Atlanta, Georgia, October 1978.

Parry MF, fiutchinson JH, Brown NA, Chii-Huei W, Estreller L: Gram-negative sepsis in neonates: a nursery outbreak due to hand carriage of Citrobacter diversus. Pediatrics 1960; 65: 1105-1109.

Gwynn CM, George RH: Neonatal citrobacter meningitis. Arch Dis Child 1973; 48: 455-458.

Riebeiro CD, Davis P, Jones DM: Citrobacter koseri menin- gitis in a special care baby unit. J Clin Pathol 1976; 29: 1094-1096.

Moorhead CD, Houck PW: Epidemiology of Pseudomonas infections in a pediatric intensive care unit. Am J Dis Child 1972: 124: 564-570.

Bobo RA, Newton EJ. Jones LF, Farmer JJ III: Nursery out- break of Pseudomonas aeruginosa: epidemiological con- clusions from five different typing methods. Appl Micro- biol 1973; 25: 414-416.

Steere AC, Mallison GF: Handwashing practices for the prevention of nosocomial infections. Ann Intern Med 1975; 63: 683-690.

American Academy of Pediatrics, Committee on Fetus and Newborn: Hospital care of newborn infants. 5th ed, Evanston: American Academy of Pediatrics. 1971.

March 1981 The American Journal of Medicine Volume 70 701