Pneumonia in pregnancy

William H. Goodnight, MD; David E. Soper, MD

Despite advances in antimicro-bial therapy and respiratorycare, pneumonia in the ante-partum period can be associ-

ated with significant maternal and neo-natal morbidity (1– 4). The estimatedprevalence of antepartum pneumoniaranges from 0.78 to 2.7 per 1000 deliver-ies (3, 4). This rate is similar to the non-pregnant population, with reported ratesof hospitalization for pneumonia of 1.51per 1,000 deliveries vs. 1.47 per 1,000nonpregnant controls (5). The onset ofpneumonia is not gestational age depen-dent, with the average diagnosis at 32wks of estimated gestational age (5). Ma-ternal physiologic adaptations to preg-nancy make pneumonia less well toler-ated during pregnancy. Pregnancyincreases the risk of maternal complica-

tions from community-acquired pneu-monia, including the need for mechani-cal ventilation in 10–20%, bacteremia in16%, and empyema in 8% of cases (1, 4).Pneumonia is a significant cause of hos-pitalization for respiratory disorders dur-ing pregnancy, accounting for 92 of 294respiratory admissions in obstetric pa-tients during influenza season (6). Respi-ratory failure due to pneumonia is thethird leading indication for intubationduring pregnancy, accounting for 12% ofintubated obstetric patients (2). Pneumo-thorax, atrial fibrillation, and pericardialtamponade complicate another 4% ofcases of antepartum pneumonia (1). Ma-ternal mortality from community-ac-quired pneumonia, though still signifi-cant, has been reduced with the use ofantibiotics from 23% to �4% (1, 6, 7),with most cases of maternal deaths oc-curring in patients with coexisting car-diopulmonary disease (1).

Pneumonia complicating pregnancycan also have adverse fetal effects. In theseries described by Madinger et al. (1),preterm labor occurred in 44% of cases ofantepartum pneumonia, with a pretermbirth rate of 36%. Munn et al. (4) also

reported a significantly increased risk forthe need for tocolysis (22% vs. 4.2% forcontrols), gestational age at delivery of�34 wks (22% vs. 7.6%, p � .12), andaverage gestational age at delivery of 36wks (38 wks in controls) in pregnanciescomplicated by pneumonia. Munn et al.(4) demonstrated that pregnancies com-plicated by pneumonia result in a signif-icantly lower average birth weight at de-livery. The relative risk for smallgestational age infants from pregnanciescomplicated by pneumonia is 1.86 (95%confidence interval, 1.01–3.45) (3, 5),with an average birth weight of 400 g lessthan controls (5). Pneumonia in preg-nancy also results in low-birth-weight ne-onates (weight at delivery, �2500 g) in33.9% of cases compared with 13.6% ofcontrols (4). The neonatal mortality ratedue to antepartum pneumonia rangesfrom 1.9% to 12%, with most mortalityattributable to complications of pretermbirth (1, 8).

Pregnancy results in significant phys-iologic changes that can have implica-tions for the maternal response to pneu-monia. An understanding of the normalcardiopulmonary adaptations to preg-

From the Division of Maternal–Fetal Medicine(WHG) and the Department of Obstetrics and Gynecol-ogy (DES), Medical University of South Carolina,Charleston, SC.

The authors have no financial interests to disclose.Copyright © 2005 by the Society of Critical Care

Medicine and Lippincott Williams & Wilkins

DOI: 10.1097/01.CCM.0000182483.24836.66

Objective: Historically, pneumonia during pregnancy has beenassociated with increased morbidity and mortality compared withnonpregnant women. The goal of this article is to review currentliterature describing pneumonia in pregnancy. This review willidentify maternal risk factors, potential complications, and pre-natal outcomes associated with pneumonia and describe thecontemporary management of the varied causes of pneumonia inpregnancy.

Results: Coexisting maternal disease, including asthma andanemia, increase the risk of contracting pneumonia in pregnancy.Neonatal effects of pneumonia in pregnancy include low birthweight and increased risk of preterm birth, and serious maternalcomplications include respiratory failure. Community-acquiredpneumonia is the most common form of pneumonia in pregnancy,with Streptococcus pneumoniae, Haemophilus influenzae, andMycoplasma pneumoniae accounting for most identified bacterialorganisms. Beta-lactam and macrolide antibiotics are consideredsafe in pregnancy and are effective for most community-acquiredpneumonia in pregnancy. Viral respiratory infections, including

varicella, influenza, and severe acute respiratory syndrome, canbe associated with maternal pneumonia. Current antiviral andrespiratory therapies can reduce maternal morbidity and mortalityfrom viral pneumonia. Influenza vaccination can reduce the prev-alence of respiratory hospitalizations among pregnant womenduring influenza season. Pneumocystis pneumonia continues tocarry significant maternal risk to an immunocompromised popu-lation. Prevention and treatment of Pneumocystis pneumonia withtrimethoprim/sulfamethoxazole is effective in reducing this risk.

Conclusions: Prompt diagnosis and treatment with contem-porary antimicrobial therapy and intensive care unit manage-ment of respiratory compromise has reduced the maternalmorbidity and mortality due to pneumonia in pregnancy. Pre-vention with vaccination in at-risk populations may reduce theprevalence and severity of pneumonia in pregnant women.(Crit Care Med 2005; 33[Suppl.]:S390 –S397)

KEY WORDS: pneumonia; pregnancy; varicella; influenza; severeacute respiratory syndrome; respiratory failure

S390 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

nancy is important to aid in recognizingthe effects of pneumonia and managingrespiratory compromise. Hormonal ef-fects of progesterone and beta-humanchorionic gonadotropins, alterations inchest shape and dimensions, and eleva-tion of the diaphragm from the pregnantuterus are purported causes of respira-tory changes during pregnancy. As a re-sult of these alterations, most patientsexperience a baseline perception of dys-pnea, peaking in the early third trimester(9). Maternal oxygen consumption in-creases by 15–20% (9). Pulmonary func-tion adaptations to pregnancy result in a30–40% increase in tidal volume (10),with an average tidal volume of 700 mL(9). The respiratory rate and vital capacityare not changed, resulting in a net 20%decrease in expiratory reserve volume,functional residual capacity, and residualvolume (10). These changes in pulmo-nary function may result in a diminishedability to compensate for the effects ofrespiratory disease during pregnancy.

Respiratory and renal adaptations topregnancy have implications for maternalacid–base status. Pregnancy results in acompensated respiratory alkalosis. Asminute ventilation increases by 30–40%,PaO2 increases to 104–108 mm Hg, andPaCO2 decreases to 27–32 mm Hg at base-line (10). The arterial pH remains in thenormal range due to increased renal ex-cretion of bicarbonate (10). Alterationsfrom these compensated values due torespiratory illness can affect fetal oxygen-ation, and minor changes in PaCO2 andPaO2 may indicate a more severe respira-tory dysfunction than clinically apparent.

Certain immunologic alterations inpregnancy may have effects on maternalsusceptibility and response to pneumonia(11). Pregnancy is associated with a de-crease in cell-mediated immunity (11,12), a decrease in helper-T-cell numbers,and a decrease in natural killer cell activ-ity (11). The increased susceptibility toviral and fungal pneumonias in preg-nancy has been attributed to the preg-nancy-induced decrease in cytotoxic lym-phocyte activity (11).

Community-Acquired Pneumonia

Pneumonia is the result of an infec-tion of the distal bronchioles and alveoli(9). Most organisms are acquired by in-halation or aspiration of nasopharyngealsecretions (9, 13). Infection causes directlung injury and interstitial inflammationmediated by the host response. This lung

injury results in intrapulmonary shunt-ing and creation of ventilation–perfusionmismatch, contributing to potential hyp-oxia (13).

In pregnancy, as in the nonpregnantpopulation, the etiological agent is notidentified in 40–61% of cases of commu-nity-acquired pneumonia (1, 7, 14). Themost common bacterial agents identifiedin pregnancy include Streptococcuspneumoniae in 17% of cases and Hae-mophilus influenzae identified in 6% ofcases (15). Viral pneumonia contributesto 5% of identified pathogens in pneumo-nia during pregnancy, with varicella andinfluenza the most common viral patho-gens (15). Other organisms identified in-clude Mycoplasma, Staphylococcus au-reus, Legionella pneumophila, Klebsiellapneumoniae, and Pseudomonas (9, 11,15–17). Fungal and protozoal organismscan also result in pulmonary infectionduring pregnancy, usually affecting im-munocompromised populations. Identi-fied risk factors for the development ofpneumonia during pregnancy includepreexisting maternal disease (HIV,asthma, cystic fibrosis), anemia, cocaineuse, and alcohol abuse (1, 3, 4, 7, 18).Madinger et al. (1) reported that 24% ofpatients with antepartum pneumonia hadan underlying maternal illness. Maternalasthma (odds ratio, 5.3) and anemia (oddsratio, 9.9) are significantly associatedwith the development of pneumonia dur-ing pregnancy (4). The use of corticoste-roids for enhancement of fetal lung ma-turity and tocolytic agents has also beenassociated with antepartum pneumonia(15). Among cases of pneumonia duringpregnancy, the pneumonia was morelikely to be hospital-acquired than com-munity-acquired among patients receiv-ing corticosteroids for fetal lung maturity(4). Tocolytic agents carry an increasedrisk for pulmonary edema that mayworsen the respiratory status of coexist-ing pneumonia or confuse the diagnosisof pneumonia. As pneumonia increasesthe risk for preterm labor, judicious useof tocolytic agents during pneumonia iswarranted.

Clinical symptoms of pneumonia in-clude fever, cough, pleuritic chest pain,rigors, chills, and dyspnea (13). Duringpregnancy, 59.3% of patients reported aproductive cough, 32.2% shortness ofbreath, and 27.1% reported pleuriticchest pain (4). Physical examination usu-ally reveals tachypnea, dullness to per-cussion, tactile and vocal fremitus, ego-phony, and use of accessory muscles of

respiration (9, 13). Auscultation maydemonstrate a pleural friction rub, in-spiratory rales, or absent breath soundsover the affected lung field (13). Physicalexamination is only 47– 69% sensitiveand 58 –75% specific for pneumonia;therefore, all cases suspicious for pneu-monia, even in pregnancy, should be con-firmed by chest radiograph (13). Munn etal. (4) demonstrated that 98% of patientswith antepartum pneumonia had positivechest radiographs, either at admission oron repeat examinations, with findings in-cluding infiltrate, atelectasis, pleural ef-fusion, pneumonitis, or pulmonaryedema. Other illnesses in the differentialdiagnosis that can present with symp-toms of pneumonia include pulmonaryembolism, cholecystitis, appendicitis, andpyelonephritis. Laboratory evaluation forsuspected pneumonia during pregnancyshould include a complete blood countand serum chemistries for liver, renal,and glucose evaluation. Evaluation ofmaternal oxygen status with pulse oxim-etry or arterial blood gas should be per-formed (14). As applicable, based on ges-tational age, fetal status should beevaluated with electronic fetal monitor-ing. Identification of the etiological agentshould be attempted with sputum Gram-negative stain and culture or urinary/serologic antigen or antibody evaluation(Table 1). Blood cultures may assist inthe identification of the etiological agent,especially in admitted patients and pa-tients with severe community-acquiredpneumonia (14). However, in most seriesof pneumonia in pregnancy, blood cul-tures have been inconsistently performedand are only rarely positive (7, 11, 18).

In the nonpregnant population, sev-eral severity assessment tools have beendeveloped to predict the course of pneu-monia, the need for hospitalization, andto predict mortality. The most commonlyused guidelines have been prepared bythe American Thoracic Society (ATS) andthe British Thoracic Society. The ATSguidelines stratify patients into fourgroups of severity based on coexistingillness (COPD, asthma, chronic renal fail-ure, alcohol abuse, diabetes, malignancy,chronic liver disease), respiratory rate of�30 breaths/min, diastolic blood pres-sure of �60 mm Hg or systolic bloodpressure of �90 mm Hg, pulse of �125beats/min, temperature of �35 or�38.3C, sepsis, confusion, white cellcount of �4 or �30 � 109/L, PaO2 of�60 mm Hg or PaCO2 of �50 mm Hg,creatinine of �1.2 mg/dl or blood urea

S391Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

nitrogen of �20 mg/dl, hemoglobin of�9 mg/dl, arterial pH of �7.35, or mul-tilobar involvement or effusion on chestradiograph (14). These criteria indicateoptimal location for treatment (e.g., out-patient, inpatient, intensive care unit[ICU]), and they suggest the pathogenesisof the pneumonia directing antibioticchoice (14). Yost et al. (18) retrospec-tively applied the ATS guidelines to 119pregnant patients with pneumonia, cor-rectly identifying all patients with a com-plicated course. Furthermore, they iden-tified that 25% of their patients wouldhave met the criteria for outpatient ther-apy (18). A simpler version has been pro-posed by the British Thoracic Society thatconsiders the presence of two of four cri-teria indicating severe illness. The BritishThoracic Society criteria include: respira-tory rate of �30 breaths/min, diastolicblood pressure of �60 mm Hg, bloodurea nitrogen of �19.1 mg/dl, and con-fusion (18). Patients with any two ofthese criteria have a 36-fold increase inmortality compared with more mild ill-ness (18). Although these criteria havebeen applied only retrospectively to a lim-ited series of pregnant patients, the ATSor British Thoracic Society guidelinesmay be applicable in predicting the needfor admission, ICU admission, and anti-biotic choice in pregnant women.

Contemporary management of pneu-monia in pregnancy includes admission,initiation of antimicrobial therapy, fetalevaluation, and maintenance of normalmaternal respiratory function. ATSguidelines recommend treatment withmacrolide antibiotic for mild illness, withaddition of beta-lactam for severe illness(Table 2) (14). Yost et al. (18) demon-strated monotherapy with erythromycinwas inadequate in only 1 of 119 patientswith pneumonia in pregnancy, withtreatment failures characterized by mul-tilobar involvement, respiratory distress,and valvular heart disease (14). Macrolideand beta-lactam antibiotics have a favor-able safety profile in pregnancy and pro-vide adequate coverage for the most com-mon organisms (19, 20). In patients at anincreased risk of hospital-acquired pneu-monia or aspiration pneumonia, the ad-dition of an aminoglycoside for coveragefor Pseudomonas and enteric Gram-negative organisms should be considered(14).

Pneumonia in pregnancy results in in-flammation and edema of the alveoli, de-creasing the number available for oxygentransport (21). Vascular flow remains

present to these affected alveolar units(21). Supplemental oxygen is required totreat the increased alveolar–arterial oxy-genation gradient that results from thisventilation/perfusion mismatch and isnecessary in the majority of pregnant pa-tients with pneumonia (3). Treatment ofreactive airway disease and chest physicaltherapy is a useful adjuvant to improverespiratory function. The increased affin-ity for oxygen by fetal hemoglobin createsan oxygen dissociation curve for fetal he-moglobin that favors transplacentaltransfer of oxygen from the mother to thefetus. This increased oxygen affinitymakes the fetus resistant to mild changesin maternal PaO2. Fetal delivery of oxygenwill decrease when the maternal oxygensaturation falls to �90%, correspondingto a PaO2 of 65 mm Hg (21). The goal oftherapy therefore should be to maintainthe maternal PaO2 of �60–70 mm Hgwith the lowest possible FIO2 to ensureadequate fetal oxygenation (21). Respira-tory failure occurs in 10% of patientswith pneumonia in pregnancy, despiteantibiotic therapy. Indications for ICU ad-mission and intubation with mechanicalventilation include inadequate oxygen-ation (PaO2 of �60 mm Hg or oxygensaturation of �85% on 0.6 FIO2), inade-quate ventilation (PaCO2 of �50 mm Hg),airway protection, sepsis requiring inva-sive hemodynamic monitoring, or persis-tent metabolic acidosis (2, 11, 21). As thepathogenesis of respiratory compromisein pneumonia is that of intrapulmonaryshunting rather that hypoventilation, theaddition of positive end-expiratory pres-sure will allow a lower FIO2 to preventcollapse of alveoli and reduce the alveo-lar-arterial oxygen gradient (21). Case re-ports of the use of high-frequency oscil-latory and positive-pressure ventilationand an intravenacaval membrane oxygen-ator for refractory cases of antepartumpneumonia have been described (11).Elective delivery has also been advocatedto improve maternal respiratory status;however, few studies have evaluated ma-ternal respiratory response to delivery. Inone case series, nine intubated pregnantpatients underwent delivery, resulting ina 28% reduction in oxygen requirementwithin 24 hrs of delivery (22). No otherchanges in ventilatory indices or clinicalcourse were identified, leading the au-thors to conclude that delivery should beperformed only for obstetric indications(22).

Fungal Pneumonia

Fungal pathogens that have been as-sociated with pneumonia include Crypto-coccus neoformans, Histoplasma capsu-latum, Sporothrix schenckii, Blastomycesdermatitidis, and Coccidioides immitis (9).These organisms are acquired from envi-ronmental sources with regional predilec-tions and usually cause mild, self-limitingdisease. Pneumonia in pregnancy with fun-gal organisms is rare. Isolated fungal pneu-monia in pregnancy usually resolves withor without treatment in women withoutcoexisting illness (9, 23). In contrast, dis-seminated disease carries a more seriousprognosis (23, 24). Twenty percent of pa-tients with coccidioidomycosis pneumoniain the third trimester of pregnancy devel-oped disseminated disease, with an in-creased risk of preterm delivery, perinatalmortality, and a high rate of maternal mor-tality (24). Disseminated disease was morecommon with infection in the third trimes-ter (24). Ely et al. (23) reported a 29%maternal mortality rate for disseminatedcryptococcal infection in pregnant, immu-nocompetent women, whereas no maternaldeaths were noted in a case series of fourpatients with isolated antepartum crypto-coccal pneumonia.

Fungal pneumonia may present withslow onset of cough and dyspnea or anacute onset of pleuritic chest pain withhypoxemia (23). Chest radiographs tendto demonstrate nodular disease and ade-nopathy, but they may show lobar ormultilobar airspace disease (23). The di-agnosis may be confirmed by sputumGram-negative stain and culture for fun-gal organisms or by the detection of se-rum fungal antigens. For disseminateddisease or severe pneumonia (ATS orBritish Thoracic Society criteria) treat-ment with intravenous amphotericin B(pregnancy category B) is recommended,followed by oral fluconazole postpartum(23, 24). Mild, isolated fungal pneumoniamay be observed with close monitoring ofchest radiographs and respiratory statusor treated with amphotericin B or flucon-azole (23). Although single-dose oral flu-conazole during pregnancy does notseem to increase the risk of congenitalmalformation (25, 26), case reports sug-gest long-term, parenteral use is associ-ated with an increased risk of fetal anom-alies, including brachycephaly, abnormalfacies, abnormal calvarial development,and cleft palate (19, 27).

S392 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

Viral Pneumonia

Varicella and influenza are the mostcommon pathogens associated with viralpneumonia in pregnancy (7, 9, 11, 28).Other viral pathogeneses have also re-sulted in pneumonia during pregnancy,including rubella, rubeola, Hantavirus,and severe acute respiratory syndrome(SARS) (9, 29–31). Viral invasion of lungparenchyma results in an interstitialpneumonitis, exacerbated by the host im-mune response to the infection (32). Theresult is significant impairment of pul-monary gas exchange, poorly tolerated bythe pregnancy-adapted respiratory sys-tem (32). Viral pneumonia is often com-plicated by acute respiratory failure, sec-ondary bacterial infections, and adultrespiratory distress syndrome (9).

Although primary varicella infection isa childhood illness, 5–10% of cases occurafter age 15 (33, 34). This uncommonadult infection, however, accounts for25–55% of fatal cases of varicella (33, 34).Approximately 10% of the adult popula-tion is susceptible to primary varicellainfection (35). The risk of primary vari-cella infection in at-risk adults after closeexposure may be as high as 70% (35).Acute varicella-zoster virus infection af-fects 0.5–0.7 of 1,000 pregnancies (33).The varicella virus is a highly contagioushuman DNA herpes virus transmitted byrespiratory droplets and close personalcontact (33). Household contact attackrates approach 90% (33). After an incu-bation period of 10 to 21 days, primaryvaricella-zoster virus infection presentswith fever, headache, and malaise, fol-lowed by the characteristic pruritic mac-ulopapular to vesicular rash. Most casesresolve within 7–10 days after the onsetof the rash (33). Complications of pri-mary varicella-zoster virus infection,more common in adults, include second-ary bacterial cellulitis, encephalitis, orpneumonia (33).

Pulmonary involvement in primaryvaricella-zoster virus infection is noted in16% of cases (33). Varicella pneumoniacomplicates 5.5–16.5% of cases of adultprimary varicella (36). Approximately 3–5days after the onset of the rash, signs andsymptoms of pneumonia may becomepresent, including a vesicular rash, orallesions, dyspnea, cough with blood-tinged sputum, malaise, and pleurisy (11,34, 37). The diagnosis of varicella pneu-monia is confirmed by the presence of aninterstitial, nodular pattern (“ground-glass” appearance) or focal infiltrates on

chest radiograph in a patient with pri-mary varicella symptoms including char-acteristic rash and fever (36). Presence ofmaternal immunoglobulin M and sero-conversion of immunoglobulin G anti-bodies from acute to convalescent phaseto varicella confirm the diagnosis but re-quires �2 wks for seroconversion. Viralculture or polymerase chain reactionidentification of varicella DNA from thebase of lesion can also confirm the diag-nosis of primary varicella. Varicellacauses an interstitial pneumonitis medi-ated by the host immune response (32).Pathologic changes in bronchioles in-clude mononuclear cell infiltrates, capil-lary endothelial cell injury, intra-alveolarexudates, and hemorrhage (32). Thesepathologic changes result in significantimpairment of pulmonary gas exchange(32).

The risk of varicella pneumonia com-plicating primary varicella-zoster virusinfection during pregnancy (0.1–18.3%)is similar to the nonpregnant state. Be-fore the introduction of antiviral therapy,the mortality rate for varicella pneumo-nia in pregnancy was significantly higher(41%) than in nonpregnant patients (1.5–12.1%) (36, 38, 39). Contemporary ma-ternal mortality remains high in mostreports, ranging from 11% to 35% (11,32, 37, 39), with one series of 18 patientswith no maternal deaths (34). Risk fac-tors for varicella pneumonia include latergestational age (39), history of or currentsmoking, and skin involvement with�100 vesicles (34). Varicella pneumoniais more likely to occur in the second orthird trimester, with average gestationalage at onset of 27 wks and average gesta-tional age at delivery of 36 wks (39). Mostpatients in modern case series are treatedwith acyclovir, and all demonstrate dif-fuse nodular densities or diffuse reticularinfiltrates on chest radiographs (37, 39).The prominent hypoxemia due to pneu-monitis results in a high rate of respira-tory failure. Mechanical ventilation maybe required in up to 40–57% of pregnantpatients with varicella pneumonia (37,39). The need for mechanical ventilationincreases the mortality rate to 25% (39).The more prevalent use of acyclovir mayreduce the risk of respiratory failure, asdemonstrated in the case series describedby Harger et al (34). Nearly all patients inthis series received acyclovir, resulting inan 11% need for mechanical ventilationand no maternal deaths (34). Acyclovirtherapy resulted in a reduction in mater-nal mortality from 36% (historical con-

trol) to 13%, and fetal mortality was re-duced from 48% to 6% in a case seriesdescribed by Broussard et al (40).

The management of maternal expo-sure to varicella during pregnancy isbased on the maternal immune status tovaricella (Fig. 1). Previous, known vari-cella infection, previous vaccination, orthe presence of serum varicella immuno-globulin G confers immunity and no ma-ternal or fetal risk from exposure. In asusceptible gravid, administration of vari-cella-zoster immunoglobulin is recom-mended within 96 hrs of exposure to pre-vent maternal illness (35, 41, 42). Theability to prevent congenital varicellasyndrome with varicella-zoster immuno-globulin is unknown. Administration oforal acyclovir (800 mg five times daily) isrecommended for pregnant women withprimary varicella infection to prevent se-rious complications such as pneumonia,but it is most effective if given within thefirst 24 hrs of the onset of the rash (35,42). Pregnant women with primary vari-cella infection who develop respiratorysymptoms should be evaluated and man-aged early for varicella pneumonia. Inthis situation, admission is likely indi-cated (35, 36, 42), and fetal and maternalevaluation for hypoxemia and respiratoryfailure should be undertaken. Intrave-nous acyclovir is recommended for clin-ically apparent varicella pneumonia andfor a susceptible mother who developsrespiratory symptoms within 10 days of aknown exposure to varicella (35, 40, 42).The acyclovir dose for varicella pneumo-nia is 10 mg/kg every 8 hrs intravenouslyfor �5 days (35, 36, 42).

Recognition and treatment of mater-nal hypoxia is important to reduce ma-ternal and fetal morbidity. Potgieter andHammond (43) reviewed 15 adult ICUpatients with varicella pneumonia. Onlythree of eight patients achieved adequateoxygenation with face-mask supplemen-tation (43). The majority of varicella pa-tients were managed with face-mask con-tinuous positive airway pressure (43).Indications for continuous positive air-way pressure included hypoxemia (PaO2

of �60 mm Hg) with adequate alveolarventilation (PaCO2 of �42 mm Hg) incooperative patients able to cough andprotect their airway (43). Four of 15 pa-tients failed continuous positive airwaypressure by face mask, requiring intuba-tion and positive-pressure ventilation(43). The effective use of face-mask con-tinuous positive airway pressure may re-duce potential complications with intu-

S393Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

bation and mechanical ventilation whilecorrecting the profound hypoxemia asso-ciated with varicella pneumonia.

The addition of corticosteroids to an-tibiotics and oxygen therapy has shownimprovement in respiratory function inserious respiratory conditions such asacute respiratory distress syndrome andviral pneumonia. Corticosteroids may re-duce the host intrapulmonary inflamma-tory response, thereby decreasing the de-gree of hypoxemia (32). The addition ofcorticosteroids to antiviral therapyamong ICU patients with varicella pneu-monia resulted in a reduction in length ofICU and hospital stays and improvementin survival (32). No patient in the steroidgroup died, whereas mortality in the con-ventional therapy group was 33% (32).Cheng et al. (38) reviewed 120 cases ofvaricella pneumonia. The mortality rate

among patients treated with antiviralagents alone was 10.3%, whereas the ad-dition of corticosteroids resulted in 100%survival in 19 patients with varicellapneumonia (38). Hydrocortisone wasused in a dose of 200 mg intravenouslyevery 6 hrs for 48 hrs (32, 38). Lee et al.(44) described the use of extracorporeallife support in seven patients with vari-cella pneumonia with respiratory failureunresponsive to medical therapy, two ofwhom developed pneumonia duringpregnancy. Indications for extracorporeallife support in this series included shuntfraction of �30%, alveolar–arterial gradi-ent of �600 torr, or PaO2/FIO2 ratio of�80, despite maximal conventional ther-apy (44). Overall survival was five of sevenpatients; one patient who developedpneumonia at 17 wks of gestational age

died, and the other peripartum patientsurvived (44).

Maternal varicella infection can havefetal effects as well. Preterm birth oc-curred in 14.3% of pregnancies with pri-mary varicella infection (45). Intrauter-ine infection by varicella can bedocumented in 24% of infants (46). Ma-ternal primary varicella infection be-tween 8 and 20 wks of gestational age canresult in development of the fetal congen-ital varicella syndrome in approximately1.2–2% of cases (45, 46). This syndromeis characterized by scarring in a derma-tomal distribution, cataracts, chorioreti-nitis, limb hypoplasia, and microcephaly(45). The onset of maternal primary vari-cella infection 5 days before to 2 daysafter delivery can result in neonatal in-fection in 17–30% of newborns, with aneonatal mortality rate of 31% (47). Useof varicella-zoster immunoglobulin re-duces the neonatal mortality rate to 7%(47). Whereas preconceptional evaluationof maternal risk to varicella and vaccina-tion before conception can prevent pri-mary varicella infection in pregnancy,early evaluation and treatment with vari-cella-zoster immunoglobulin and acyclo-vir may prevent serious complications ofmaternal disease. Treatment of varicellapneumonia with acyclovir and correctionof respiratory failure can reduce maternalmortality.

Influenza A and B are common world-wide causes of respiratory illness; influ-enza A is the most virulent strain in hu-mans. The influenza virus is classified byfour hemagglutinin and two neuramini-dase antigen subtypes. Major antigenchanges, known as shifts, occur slowlyand give rise to epidemics (48). Minorantigen changes or drifts occur more fre-quently and are classified by year andlocation of identification (48). Influenzavaccine preparation is based on predicteddrifts for the upcoming season deter-mined by worldwide influenza activity(48, 49). Influenza infection is usually aself-limited infection, characterized bymalaise, fever, myalgia, cough, rhinor-rhea, and nausea/vomiting (9, 11, 50, 51).

Pregnancy increases the risk of com-plications from influenza (50). During in-fluenza epidemics of 1918 and 1957, mor-tality from influenza during pregnancyreached 30 –50% (11, 50). Pregnantwomen during influenza season are af-fected more frequently than nonpregnantwomen, with influenza-related morbidityoccurring in 10.5 of 10,000 (95% confi-dence interval, 6.7–14.3) pregnant

Figure 1. Prophylaxis for maternal varicella during pregnancy. IgG, immunoglobulin G; VZIG,varicella-zoster immunoglobulin.

S394 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

women, compared with a rate of 1.91 of10,000 (95% confidence interval, 1.51–2.31) nonpregnant controls (52). Thisrate is similar to nonpregnant patientswith coexisting illness (52). Excess ratesof influenza, pneumonia, upper respira-tory infection, and respiratory symptomsare experienced by pregnant women com-pared with a nonpregnant population, es-pecially in relation to new antigenic shiftsin viral activity (53).

Influenza pneumonia occurred in 12%of patients in a series of 102 pregnantpatients with influenza during the 2003–2004 season (51). One patient requiredintubation and mechanical ventilation(8%), and there were no maternal deaths(51). Other complications included men-ingitis (1%) and myocarditis (1%). Influ-enza pneumonia carries significant mor-tality in pregnant and nonpregnantpatients from respiratory failure, withmortality ranging from 12.5% to 42.1%(28, 50). Antiviral medications are effec-tive in preventing and treating influenzaillness. Amantadine and rimantadine areeffective against influenza A, with pro-phylactic use preventing 70–90% of ill-nesses among exposed patients (49). Ad-ministered within the first 48 hrs of theonset of illness, amantadine and riman-tadine are effective in shortening thecourse of symptoms and decreasing viralload present in secretions (49). Case re-ports of amantadine use in pregnancyhave been reassuring in its safety profile,with rare reports of cardiac defects afterfirst-trimester exposure slightly abovethe expected population rate (18, 54, 55).Newer antiviral medications include theneuraminidase inhibitors zanamivir andoseltamivir. These agents are indicatedfor prophylaxis and treatment for bothinfluenza A and B if administered withinthe first 48 hrs of the onset of illness (49).Complications such as sinusitis, bronchi-tis, and otitis media are reduced with theuse of neuraminidase inhibitors (49).There are no studies of the use of neur-aminidase inhibitors in pregnancy (19).The use of antiviral medications can re-duce the mortality from influenza pneu-monia from 42.1% in conservativelymanaged patients to 27.3% (38). As invaricella pneumonia, the addition of cor-ticosteroids to conservative treatmentalso demonstrates reduced mortalityfrom influenza pneumonia to a rate of12.5% in a limited case series (38).

SARS is a new viral illness first de-scribed in 2002. SARS results in an atyp-ical pneumonia caused by a previously

undescribed coronavirus that can rapidlyprogress to respiratory failure (31).Symptoms usually develop 2–7 days afterexposure and include fever, chills, rigors,headache, malaise, and myalgia (56). Anonproductive cough or dyspnea devel-ops over 3–7 days. This respiratory phasecan progress to hypoxemia and respira-tory failure in 10–20% or cases (56).Chest radiographs demonstrate general-ized, patchy, interstitial infiltrates, andlaboratory evaluation can demonstrateleukopenia, thrombocytopenia, elevatedcreatine kinase, and elevated hepatictransaminases (56). Confirmation of thediagnosis is made by polymerase chainreaction for SARS virus on two differentspecimens, seroconversion by enzyme-linked immunosorbent assay, or viral iso-lation by culture (56, 57). The overallmortality rate for SARS is 3%, whereasserious illness resulting in ICU admissioncarries a 20% mortality rate (56).

Wong et al. (31) reviewed 12 cases ofSARS occurring during pregnancy, sevenduring the first trimester and five occur-ring in the second or third trimesters.Fifty percent of the patients were admit-ted to the ICU, 33% requiring mechanicalventilation for a range of 16–37 days (31).The maternal mortality rate was 25% inthis small sample, compared with anoverall case fatality rate of 3% (31, 56).Neonatal morbidity was high, with four ofthe seven patients presenting in the firsttrimester complicated by spontaneousabortion (31). Of the five pregnanciespresenting after 26 wks, three were deliv-ered preterm for maternal or fetal indi-cations. Of the two remaining pregnan-cies, one delivered at 33 wks and one at37 wks, both complicated by a small forgestational age fetus and placenta weightless than the fifth percentile (31). Placen-tal pathology demonstrated avascular villiand placental infarcts, presumed second-ary to the maternal hypoxemia and circu-latory insufficiency of SARS (31). NoSARS virus was detected in any fetus orplacenta (31). All patients were treatedwith antibiotics, ribavirin, and corticoste-roids. A review of SARS cases by Cheng etal. (38) demonstrated a reduction in mor-tality in nonpregnant SARS patients withtreatment with ribavirin, oseltamivir, ste-roids, or combination steroids and anti-viral medications. Mortality with conser-vative therapy was 15.4%, whereas theuse of antiviral medications, with orwithout corticosteroids, resulted in amortality rate of 2–7% (38). As with otherviral respiratory infections, SARS carries

an increased risk of maternal and fetalmorbidity and mortality during preg-nancy.

Pneumonia inImmunocompromised Patients

Certain pathogens can cause pneumo-nia in immunocompromised patients.These include bacteria (Staphylococcus,Mycoplasma, and Mycobacterium), fun-gal, viral, and parasitic organisms (in-cluding Pneumocystis carinii) (9). Pneu-mocystis pneumonia is the mostcommon cause of AIDS-related deathamong pregnant patients (58). Most pa-tients present with dry cough, tachypnea,and dyspnea, and chest radiographs dem-onstrate diffuse interstitial infiltrates. Inmost cases, the diagnosis can be made byhistiologic staining of sputum, althoughbronchoscopy may be necessary in somecases. Pneumocystis pneumonia in preg-nancy carries significant risk to themother and fetus. In a review of 22 casesof Pneumocystis pneumonia in preg-nancy, Ahmad et al. (58) demonstrated a59% rate of respiratory failure with needfor mechanical ventilation. The overallmortality in this series was 50%, com-pared with 1–16% in a nonpregnant pop-ulation (58). Fetal mortality was high,with five intrauterine deaths and fourneonatal deaths (58). Although the pa-tients in this series were treated withtrimethoprim/sulfamethoxazole (TMP/SMX), with or without corticosteroids,the episode of Pneumocystis pneumoniawas the presenting illness of HIV/AIDSfor these patients (58). The current rec-ommendation for treatment of Pneumo-cystis pneumonia is TMP/SMX (20, 59).For patients with PaO2 of �70 mm Hg,oral TMP/SMX at a dose of two double-strength tablets every 8 hrs or 15mg·kg�1·day�1 (TMP component) every 8hrs intravenously for 21 days is recom-mended (20, 59). For severe cases, withPaO2 of �70 mm Hg or alveolar–arterialoxygen gradient of �35 mm Hg, oralprednisone or intravenous methylpred-nisolone are initiated before addition ofTMP/SMX (20, 59). Early diagnosis of HIVinfection and the use of antiretroviraltherapy can reduce the risk of severePneumocystis pneumonia in pregnancy.Current recommendations for prophy-laxis against pulmonary opportunistic in-fections in HIV include TMP/SMX dailyfor patients with previous Pneumocystisinfection or CD4 lymphocyte count of�200/�L (60). Alternative medications

S395Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

include dapsone or aerosolized pentami-dine (60). Prophylaxis for Mycobacteriumavium-intracellulare is recommended forpatients with CD4 counts of �50/�L,with 1200 mg azithromycin weekly, andyearly influenza and pneumococcal vac-cinations are recommended (60).

Prevention of Pneumonia

Several strategies are effective in pre-venting pneumonia in high-risk popula-tions and can be applied to women ofchildbearing age or during pregnancy.Vaccinations are available for influenza,pneumococcus, and varicella. The influ-enza vaccine is an inactivated virus, cre-ated annually to account for yearly anti-genic drift (48). The vaccine is effective inreducing influenza complications, in-cluding pneumonia and death, and in de-creasing physician office visits and daysmissed from work among vaccinated pop-ulations (48, 49). Vaccination against in-fluenza is able to prevent influenza illnessin 70–90% of healthy adults �65 yrs ofage during influenza season (48, 49). Alive, attenuated intranasal vaccine is alsoavailable. No adverse fetal outcomes havebeen identified in women who receivedthe inactivated vaccine during pregnancy(19). The risk for influenza-related respi-ratory illness in pregnancy is similar tohigh-risk nonpregnant populations, andpotential morbidity from influenza is in-creased in pregnancy (8, 48). Therefore,the influenza vaccine is recommended forall women who will be pregnant duringinfluenza season, regardless of gesta-tional age (48).

Evaluation of maternal risk for vari-cella, including documentation of knownvaricella infection or detection of serumvaricella immunoglobulin G, should beascertained during preconceptional eval-uation or in early pregnancy. A live atten-uated vaccine (Varivax) became availablein 1995 (47). Varicella vaccination is rec-ommended for susceptible women con-sidering pregnancy at 1–3 months beforepregnancy or postpartum (47). Vaccina-tion may reduce the risk of congenitalvaricella syndrome and decrease morbid-ity from adult complications of varicella(47). The varicella vaccine is not recom-mended for use during pregnancy.

The pneumococcal vaccine (Pneumo-vax, Pnu-Imune) is composed of purifiedcapsular polysaccharide antigens fromthe 23 clinically relevant pneumococcaltypes and is effective in decreasing theprevalence of pneumococcal pneumonia

in high-risk populations (61). The pneu-mococcal vaccine is recommended towomen with underlying medical ill-nesses, including immunocompromisedstates, asplenia, sickle cell disease, diabe-tes, or chronic cardiopulmonary disease(61). The pneumococcal vaccine carrieslittle biological suspicion for fetal effect;therefore, it may be given during preg-nancy in women with the listed risk fac-tors (19).

Several strategies may be employed toreduce the risk of aspiration pneumoniaduring pregnancy. Pregnancy results insmooth-muscle relaxation of the gastro-intestinal tract, delaying gastric empty-ing and decreasing the tone of the gas-troesophageal sphincter. The use ofcricoid pressure during intubation at-tempts, premedication with oral antacids(Bicitra) before anesthesia, and carefuluse of sedation in pregnant women hasbeen suggested to decrease the risk ofaspiration (9).

REFERENCES

1. Madinger NE, Greenspoon JS, Ellrodt AG:Pneumonia during pregnancy: Has moderntechnology improved maternal and fetal out-come? Am J Obstet Gynecol 1989; 161:657–62

2. Jenkins TM, Troiano NH, Graves CR, et al:Mechanical ventilation in an obstetric popu-lation: Characteristics and delivery rates.Am J Obstet Gynecol 2003; 188:549–552

3. Berkowitz K, LaSala A: Risk factors associ-ated with the increasing prevalence of pneu-monia during pregnancy. Am J Obstet Gy-necol 1990; 163:981–985

4. Munn MB, Groome LJ, Atterbury JL, et al:Pneumonia as a complication of pregnancy. JMatern Fetal Med 1999; 8:151–4

5. Jin Y, Carriere KC, Marrie TJ, et al: Theeffects of community-acquired pneumoniaduring pregnancy ending in a live birth. Am JObstet Gynecol 2003; 18:800–806

6. Richey SD, Roberts SW, Ramin KD, et al:Pneumonia complicating pregnancy. ObstetGynecol 1994; 84(4 Pt 1):525

7. Panting-Kemp A, Geller SE, Nguyen T, et al:Maternal deaths in an urban perinatal net-work, 1992–1998. Am J Obstet Gynecol 2000;183:1207–1212

8. Hartert TV, Neuzil KM, Shintani AK, et al:Maternal morbidity and perinatal outcomesamong pregnant women with respiratoryhospitalization during influenza season.Am J Obstet Gynecol 2003; 189:1705–1712

9. Ramsey PS, Ramin KD: Pneumonia in preg-nancy: Medical complications of pregnancy.Obstet Gynecol Clin North Am 2001; 28:553–569

10. Cruikshank DP, Wigton TR, Hays PM: Mater-nal physiology in pregnancy. In: Obstetrics:

Normal and Problem Pregnancies. Gabbe SG,Niebyl JR, Simpson JL (Eds). New York,Churchill Livingston, 1996, pp 93–95

11. Goodrum LA: Pneumonia in pregnancy. Se-min Perinatol 1997; 21:276–283

12. Ie S, Rubio ER, Alper B, et al: Respiratorycomplications of pregnancy. Obstet GynecolSurv 2001; 57:39–46

13. Kasper DL, Braunwald E, Fauci AS, et al(Eds): Harrison’s Principles of Internal Med-icine. 16th Edition. New York, McGraw-Hill,2005

14. Guidelines for the Management of Adultswith Community-acquired Pneumonia. Am JRespir Crit Care Med 2001; 163:1730–1754

15. Lim WS, Macfarlane JT, Colthorpe CL: Treat-ment of community-acquired lower respira-tory tract infections during pregnancy. Am JRespir Med 2003; 2:221–233

16. Eisenberg VH, Eidelman LA, Arbel R, et al:Legionnaire’s disease during pregnancy: Acase presentation and review of the litera-ture. Eur J Obstet Gynecol Reprod Biol 1997;72:15–18

17. Soper DE, Melone PJ, Conover WB: Legion-naire disease complicating pregnancy. ObstetGynecol 1986; 67:10S–12S

18. Yost NP, Bloom SL, Richey SD, et al: Anappraisal of treatment guidelines for antepar-tum community acquired pneumonia. Am JObstet Gynecol. 2000; 183:131–135

19. Weiner CP, Buhimschi C: Drugs for Pregnantand Lactating Women. New York, Elsevier,2005

20. Gilbert DN, Moellering RC, Eliopoulos GM,et al (Eds): The Sanford Guide to Antimicro-bial Therapy 2004. 34th Edition. Hyde Park,VT, Antimicrobial Therapy, 2004

21. Maccato M: Respiratory insufficiency due topneumonia in pregnancy. Obstet GynecolClin North Am 1991; 18:289–299

22. Tomlinson MW, Caruthers TJ, Whitty JE, etal: Does delivery improve maternal conditionin the respiratory-compromised gravida? Ob-stet Gynecol 1998; 91:108–111

23. Ely EW, Peacock JE, Haponik EF, et al: Cryp-tococcal pneumonia complicating pregnancyMedicine 1998; 77:153–167

24. Stevens DA: Coccidioidomycosis. N EnglJ Med 1995; 332:1077

25. Jick SS: Pregnancy outcomes after maternalexposure to fluconazole. Pharmacotherapy1999; 19:221–222

26. Mastroiacovo P, Mazzone T, Botto LD, et al:Prospective assessment of pregnancy out-comes after first-trimester exposure to flu-conazole. Am J Obstet Gynecol 1996; 175:1645–1650

27. Pursley TJ, Blomquist IK, Abraham J, et al:Fluconazole-induced congenital anomaliesin three infants. Clin Infect Dis 1996; 22:336–340

28. Riley L: Pneumonia and tuberculosis in preg-nancy. Infect Dis Clin North Am 1997; 11:119–134

29. Stein SJ, Greenspoon JS. Rubeola duringpregnancy. Obstet Gynecol 1991; 78(5 Pt 2):925–929

S396 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

30. Gilson GJ, Maciulla JA, Nevils BG, et al: Han-tavirus pulmonary syndrome complicatingpregnancy. Am J Obstet Gynecol 1994; 171:550–554

31. Wong SF, Chow KM, Leung TN, et al: Preg-nancy and perinatal outcomes of womenwith severe acute respiratory syndrome.Am J Obstet Gynecol 2004; 191:292–297

32. Mer M, Richards GA: Corticosteroids in life-threatening varicella pneumonia. Chest1998; 114:426–431

33. Chapman SJ: Varicella in pregnancy. SeminPerinatol 1998; 22:339–346

34. Harger JH, Ernest JM, Thurnau GR, et al:Risk factor and outcome of varicella-zosterpneumonia in pregnant women. J Infect Dis2002; 185:422–427

35. Wilkins EGL, Leen CL, McKendrick MW, etal: Management of chickenpox in the adult: Areview prepared for the UK Advisory Groupon Chickenpox on behalf of the British Soci-ety for the Study of Infection. J Infect 1998;36(Suppl 1):49–58

36. Baren JM, Henneman PL, Lewis RJ: Primaryvaricella in adults: Pneumonia, pregnancy,and hospital admission. Ann Emerg Med1996; 28:165–169

37. Cox SM, Cunningham FG, Luby J: Manage-ment of varicella pneumonia complicatingpregnancy. Am J Perinatol 1990; 7:300–301

38. Cheng VCC, Tang BSF, Wu AKL, et al: Med-ical treatment of viral pneumonia includingSARS in immunocompetent adult. J Infect2004; 49:262–273

39. Smego RA, Asperilla MO: Use of acyclovir forvaricella pneumonia during pregnancy. Ob-stet Gynecol 1991; 78:1112

40. Broussard RC, Payne DK, George RB: Treat-ment with acyclovir of varicella pneumoniain pregnancy. Chest 1991; 99:1045–47

41. Recommendations of the Advisory Commit-tee on Immunization Practice (ACIP): Vari-

cella-zoster immune globulin for the preven-tion of chickenpox. MMWR Recomm Rep1984; 33: 84–90, 95–100

42. Nathwani D, Maclean A, Conway S, et al:Varicella infections in pregnancy and thenewborn. J Infect 1998; 36(Suppl 1):59–71

43. Potgieter PD, Hammond JMJ: Intensive caremanagement of varicella pneumonia. RespirMed 1997; 91:207–212

44. Lee WA, Kolla S, Schreiner RJ, et al: Pro-longed extracorporeal life support (ECLS) forvaricella pneumonia. Critical Care Medicine1997; 25:977–982

45. Pastuszak AL, Levy M, Schick B, et al: Out-come after maternal varicella infection in thefirst 20 weeks of pregnancy. N Engl J Med1994; 330:901–905

46. Paryani SG, Arvin AM: Intrauterine infectionwith varicella-zoster virus after maternal in-fection. N Engl J Med 1986; 314:1542–1546

47. Prevention of varicella: Recommendations ofthe Advisory Committee on ImmunizationPractices (ACIP). Centers for Disease Controland Prevention. MMWR Recomm Rep 1996;45(RR-11):1–36

48. Harper SA, Fukuda K, Uyeki TM, et al: Pre-vention and control of influenza: Recom-mendations of the Advisory Committee onImmunization Practices (ACIP). MMWRRecomm Rep 2004; 53(RR-6):1–40

49. Couch RB: Prevention and treatment of in-fluenza. N Engl J Med 2000; 343:1778–1787

50. Kort BA, Cefalo RC, Baker VV: Fatal influ-enza A pneumonia in pregnancy. Am J Peri-natol 1986; 3:179–182

51. Laibl V, Sheffield J, Roberts S, et al: Presen-tation of influenza A in pregnancy during the2003–2004 influenza season. Abstr. Am J Ob-stet Gynecol 2005; 105(Suppl):S28

52. Neuzil KM, Reed GW, Mitchel EF, et al: Im-pact of influenza on acute cardiopulmonary

hospitalizations in pregnant women. Am JEpidemiol 1998; 148:1094–1102

53. Mullooly JP, Barker WH, Nolan TF Jr: Risk ofacute respiratory disease among pregnantwomen during influenza A epidemics. PublicHealth Rep 1986; 101:205–211

54. Levy M, Pastuszak A, Koren G: Fetal outcomefollowing intrauterine amantadine exposure.Reprod Toxicol 1991; 5:79–81

55. Rosa F: Amantadine pregnancy experience.Reprod Toxicol 1994; 8:531

56. Case definitions for surveillance of severeacute respiratory syndrome (SARS). WorldHealth Organization Web site. Available at:http://www.who.int/csr/sars/casedefinition.Accessed August 16, 2005

57. Preliminary Clinical Description of SevereAcute Respiratory Syndrome. World HealthOrganization Web site. Available at: http://www.who.int/csr/sars/clinical. Accessed Au-gust 16, 2005

58. Ahmad H, Mehta NJ, Manikal VM, et al:Pneumocystis carinii pneumonia in preg-nancy. Chest 2001; 120:666–671

59. Benson CA, Kaplan JE, Masur H, et al: Treat-ing opportunistic infections among HIV-exposed and infected children: Recommen-dations from CDC, the National Institutes ofHealth, and the Infectious Diseases Societyof America. MMWR Recomm Rep 2004;53(RR-15):1–112

60. Kaplan JE, Masur H, Holmes KK: Guidelinesfor preventing opportunistic infectionsamong HIV-infected persons–2002: Recom-mendations of the U.S. Public Health Serviceand the Infectious Diseases Society of Amer-ica. MMWR Recomm Rep 2002; 51(RR-8):1–52

61. Prevention of pneumococcal disease: Recom-mendations of the Advisory Committee onImmunization Practices (ACIP). MMWRRecomm Rep 1997; 46(RR-8):1–24

S397Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)