Extracorporeal membrane oxygenation and neonatal respiratory failure: Experience from the extracorporeal life support organization

Extracorporeal membrane oxygenation and neonatal respiratory failure: Experience from the extracorporeal life support organization

Extracorporeal Membrane Oxygenation and Neonatal Respiratory Failure: Experience From the Extracorporeal Life Support Organization By Charles J.H. St...

930KB Sizes 1 Downloads 48 Views

Extracorporeal Membrane Oxygenation and Neonatal Respiratory Failure: Experience From the Extracorporeal Life Support Organization By Charles

J.H. Stolar,

Sandy

M. Snedecor,

and Robert

H. Bartlett

Ann Arbor, Michigan 0 Extracorporeal membrane oxygenation (ECMO) has rescued moribund infants with respiratory failure from a variety of causes. We report the experience from 58 United States and 7 overseas ECMO centers between 1980 and 1989. Voluntarily submitted data forms provided details of diagnosis, clinical condition, ECMO indications, morbidity, and mortality. Df 3,528 infants with a predicted mortality > 80% treated with ECMO, 83% survived. Entry diagnoses and aggregate survival were: meconium aspiration syndrome (MAS) 1,358 (93%). persistent pulmonary hypertension of the newborn (PPHN) 480 (83%); congenital diaphragmatic hernia (CDH) 585 (62%); hyaline membrane disease (HMD) 532 (84%); sepsis 416 (77%); and other 185 (77%). ECMO indications were a-Ado, > 600 for 6 to 8 hours (22%), oxygenation index >40 for 4 hours (18%). acute deterioration (14%). maximal therapy failure (34%), and barotrauma (1%). Annual survival improved over 9 years except for CDH, which decreased from 70% (1987) to 56% (1989)P < .Ol). Survivors differed from non-survivors (P < .05) by birth weight (> 2 kg), gestational age (>37 weeks), entry diagnosis (MA8 PPHN, HMD, sepsis v CDH), inborn versus outborn, preECMO pH. and ECMO duration. Technical complications in 25% of patients and medical complications in 75% adversely affected survival. Annual sepsis survival improved to 75% (1989) but had significantly greater complication rates (P < .05) than other diagnoses. Multicenter data yield information not available from single institution experience. Although entry criteria and conventional therapy continue to evolve, ECMO currently improves survival from an estimated 20% to 83% overall. Individual prognosis depends on entry diagnosis, clinical condition, and complications. Cop yrighf o 199 1 by W. B. Saunders Company INDEX

WORDS:

Extracorporeal

(ECMO); neonatal respiratory

membrane

oxygenation

failure.

A

LTHOUGH EXTRACORPOREAL membrane oxygenation (ECMO) has been used since 1975,‘,’ systematic collection of information concerning its use was not begun until 1980. In its original format as the Neonatal ECMO Registry, information was collected concerning patient demographics, pre-ECMO clinical features, ECMO indications, medical and technical complications, and short-term outcome. Between 1980 and 1987,715 patients were registered from a small number of ECMO centers.3 Eighty-one percent of treated infants survived despite a predicted 80% mortality. The most frequent indications were persistent pulmonary hypertension of the newborn (PPHN), meconium aspiration syndrome (MAS), and congenital diaphragmatic hernia (CDH). Less common were neonatal sepsis and respiratory distress syndrome (RDS). Technical complications occurred in 23% and physiologic complications occurred in Journal of Pediatric Surgery, Vol26, No 5

(May), 1991:

pp 563-571

65%. Survival rates for the first 10 patients at any one center were significantly worse than for the subsequent 10 patients (74% v 84%, P < .Ol), reflecting a learning curve. Since that report, the number of ECMO centers and infants treated has increased. This report will review the demographics, clinical features, and shortterm outcome of the infants in what is now called the Neonatal ECMO Registry of the Extracorporeal Life Support Organization (incorporated 1989, Ann Arbor, MI). We speculate that if a new therapy gains acceptance and is refined, its effect will be reflected by changing trends in diagnosis, selection criteria, outcome, complications, and diagnosis-related specific features. Because ECMO is specifically for moribund infants refractory to conventional therapy, standard randomized clinical trials have been problematic. Innovative study design and statistical methods,“.5 although controversial, suggest many advantages for ECMO compared with conventional therapy. Although Registry data are not trials of one therapy versus another, we hypothesized that if ECMO is a successful method of treatment, then cumulative and annual data should show a steady increase in survival, with decreasing morbidity and mortality despite any learning curve effects. Further, morbidity and mortality outcome for a specific ECMO entry diagnosis that differed significantly from the aggregate total might reflect pathophysiologic problems unique to that diagnosis rather than an inherent weakness in the ECMO therapy. MATERIALS

AND

METHODS

Registry data were collected from 58 domestic and 7 international ECMO centers (appendix). Data were abstracted from patient records and ECMO flow charts and submitted at the

From The Extracorporeal Life Support Organization, Ann Arbor, Ml. Presented at the 37th Annual International Congress of the British Association of Paediatric Surgeons, Glasgow. Scotland, July 25-27, 1990. Supported in part by The William Randolph Hearst Foundation, Inc, New York, NY,; The Charles Edison Fund, Orange, NJ: and The Anya Fund, Annonk, NY Address reprint requests to Charles J.H. Stolar. MD, Division of Pediatric Surgery The Babies Hospital, Room 203N, 3959 Broadway, New York, NY 10032. Copyright o 1991 by WB. Saunders Company 0022-346819112605-0015$03.00/0 563

STOLAR, SNEDECOR, AND BARTLETT

564

ECMO Center:

Date Completed: -I-

Form Completed by:

Phone:(

i-

)

Patient ldentihcation Hosprtal ID*:

Name: Sex: 1-I

M

/_I

F

Comptiition

Birthdate: _

* 1-l-J

Birthweight: 1-I

/ _

/ _

Mom?

/_I

Yes

/_I

No

Gsstatrcnal Age: I-1-I

Apgar Score: I_/_/

kg.

(weeks)

(1 mm) ~-1-l

(5 min)

of pregnancy (Describe): Pre-ECMO Course Secondary Respiratory Diagnose (Circle as many as necessary) RDS CDH Sepsis (Pneumonia) MAS PPHN (PFC) Cardrac Pneumothorax other

Primary Respiratory Diagnosis (clrde one only) MAS RDS Sepsis (Pneumonra) CDH PPHN (PFC) Pneumothcfax Cardiac other: Non-Respiratory Pre-ECMO

or Congenital Anomolies:

serzures: I-1

Yes

/_I

Renal Farlure: 1-I (Creatrnrne 2 1.5)

No

Yes

/_I

No

Other Abnormakttes: Vasoactrve Drugs Trted: I-1

None

I-I

Dopamrne

/_I

Tolazotine

l-1

Other

Last

Arteriaf Blood Gas Pnoc to ECMO: pH l-1 * l_l_l PO2 ~-l_I_~ mmHg Ventilator Setting Prior to ECMO: Fi021-I. l_l_l Rate l_/_l_l Jet/Oscrllatory Ventilations

1-I

PC02 /_l_/_I

Peak Pressure l-I-1

Yes

I-1

HCOJ l-l-l-l

mmlig

PEEP I-I-

No

Criteria for going on ECMO (check only one): 1-I AaCQ > /_/_l_t for l_l_l hours 1-l Oxygenation Index Score /_I_1 (Fi02 x MAP/PO*) 1-1 Acute Deterioration: PaO, < l_/_L mmHg for l_l_l hours Elevated MAP Emphysema 1-I Barotrauma: (circle one) PIE PT I-1 Failure to respond to maximum treatment l-l Cardiac Arrest

1-I

other: ECMO Course

Age starting ECMO: ~_~__I--I ~cde:

I_/

V-A

Cannula srze: PDA ligation whrle on ECMO:I_I

1-I

V-V

(hours of kfe) 1-l

VV-VA

Venous l-l-I

Yes

I_/

French

No

Hours on ECMO:

l-l-l-l

hours

(reason) Arterial ~-1-I

Other Surgery: /_I Fig 1.

French

Yes

I-1

No

(Describe)

Neonatal ECMO Registry forms.

conclusion of the infant’s hospitalization using data capture forms (Figs 1A and 1B). Although all infants requiring ECMO had pulmonary hypertension to some degree, a central primary diagnosis was assigned to each patient, namely MAS, CDH, neonatal sepsis, RDS, or other. RDS included classical hyaline membrane disease, perinatal asphyxia, “shock lung,” cardiac arrest, maternal drug dependency, and miscellaneous conditions. If the etiology of the pulmonary hypertension was truly idiopathic the diagnosis of PPHN was assigned. A secondary associated diagnosis of MAS, PPHN, sepsis or air leak syndrome was assigned if appropriate. Each participating center developed ECMO selection criteria estimated to represent a predicted mortality of at least 80% using maximal conventional therapy at that institution. Details of specific therapy were not a defined part of the Registry information: most institutions practiced a respiratory therapy strategy based on muscle paralysis and induced alkalosis. Failure of maximal therapy

was determined by clinical judgment, duration of alveolar-arterial oxygen gradient >600 mm Hg beyond 6 to 8 hours ([a-Ado, = [FiO, x (760 to 47 mm Hg) - [pCOJ.S) - pa02]) or oxygenation index >40 beyond 4 hours (01 = [FiO, x MAP/paO,] x 100. Other infants were entered because of acute deterioration before meeting the defined time requirements of a-Ado, or 01. ECMO therapy was administered as venoarterial bypass. Less than 2% were supported by venovenous bypass. Extrathoracic cannulation was accomplished through the right neck. The extracorporeal circuit has been previously described.6 The details of ECMO patient management are well described elsewhere.’ Physiological complications are defined on the ECMO Registry form (Figs 1A and 1B) and are characterized as either infectious, hematologic, renal, neurological, or cardiopulmonary. Technical complications are compiled on an occurrence basis and are also listed on the Registry

Form.

REGISTRY REPORT OF NEONATAL ECMO

565

Ventilator setting at 24 hours on ECMO: Rate l-l-1 FiO2 I-1. I-1-1

Peak Ressure I-1-1

PEEP I-I-I

Compkcations on ECMO (see codes below): Mechamcal:

A

B

C

D

E

F

Pabent.

1 13 25

2 14 26

3 15 27

4 16 28

5 17 29

6 18 30

7 19 31

8 20 32

9 21 33

10 22 34

11 23 35

12 24 36

Descriptions:

I-1

outcome

SUrvlVed

Hours to Extubatlon /_I_/-_ Date of Discharge Home _ Total Hospital Days l_l_/_l Pre-Discharge Morbidity (Describe)

Late Death: _

/ _

-----_---_I_

I-1 /_

hours / _

Died

Date of Death: _ I_ / ECMO Electivefy Terminated /_I Cause(s) of Death:

Yes

/_I

No

f_

----------

Mechanica/ A. oxygenator failure 0. tubing rupture, circuit disrupbon C. pump malfunction D. heat exchanger malfunctii E. cannula placement or removal problems (specify above) F. other (specify above)

Hemorrhagfc -&nonWqeby 1. Intracranial cranial ultrasound (specify) 2. lntracranlal hemorrhage by CT scan (specify above) 3. sgnifkzant GI hemonbags 4. sgnlficant surgical site bleeding 5. 25 units platelet transfuuons/ 24 hours 6. serumhgbr100 7 other (specify above)

Codes for Complitiis

Neurologic 8. brain death (no spent movement,pupils fixed and dilated) 9. probable or definite seozure 10. excessive jitteriness 11. other (specify above)

Renal 12. 13. 14. 15.

~-

on ECMO

aeatinine > 1.5. ~3.0 creatinine 2 3.0 dialysis or hemofiltratii other (specify above)

Cardiovascular 16. cardiopulmonary resuscitation required 17. cardiac arrhythmia 10. tolazoline (cu other puWxmaq vasodilator) given 19. other (specify above)

infectious 22. culture proven mfecbon (specifv _ above) 23. WBCS1500 24. other (specify above)

Metabolic 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36

K (_ 2.5 K 2 7.5 Na 5 120 Naz160 Ca < 6.0 Ca t 14.0 gkmm? 5 20 glucose 2 350 pti 5 7.05 pH t 7.75 Hypertensive: Systdi BP z 90 mmHg for 4 hours Other (specify above)

fulmonary 20. pneumothorax requiring chest tube @acement 21. other (specify above)

Return Form To: ECMO Data Registry Phone: (313) 9365822 2920 Taubman Center University of Michigan Medical Center Ann Arbor, Michigan 48109-0331 ___---_--

(Rev. l/29/87) Fig 1. (Cont’d).

Survival was defined as being decannulated from ECMO and extubated from all mechanical ventilator support for at least 24 hours.

Statistical

tics by blood gas values and ventilator settings. Logistic regression analysis was used for diagnosis specific medical complications. All statistical instruments were applied within the Statistical Analysis System @AS) and SYSTAT.

Analysis

Registry data were analysed in two ways. The xZ test (MantelHaenzel) examined the association between outcome and preECMO characteristics, technical, and medical complications. The two-sample r test identified survivor and nonsurvivor characteris-

RESULTS

Between 1980 and 1989, 3,528 infants reported from 58 centers in the United States and 7 interna-

566

STOLAR,

tional centers were registered in the Neonatal ECMO Registry of the Extracorporeal Life Support Organization. This represents a further 2,813 patients from an additional 47 centers since the first Registry report of 715 patients from 18 centers.3 During the last 3 years, 1987 through 1989, there was a 261% increase in the number of ECMO centers and a 393% increase in treated infants (Figs 2 and 3). Although there was a rapid increase in the aggregate and annual numbers of infants treated through 1988, the annual totals for 1988 and 1989 are the same. Similarly, the annual rate at which new ECMO centers have opened was greatest in 1987 and is less for 1988 and 1989 (Fig 2). The number of infants treated annually at any one ECMO center ranged from 2 to 53, with an average of 20 ECMO patients per center per year. Ninety-three percent of the patients were outborn and only 7% were inborn, indicating that almost all infants were referred to a level III neonatal intensive care unit specifically for ECMO therapy. ECMO Patient Description The primary diagnoses for the aggregate group were MAS (39%), CDH (16%), hyaline membrane disease (15%), PPHN (13%) sepsis (12%), cardiac support (2%), and other (2%). Figure 4 shows the annual total of cases by primary diagnosis. Although MAS predominates in all years, it has decreased from 47% of all ECMO cases in 1985 to 37% in 1989 (P < .05). Over the same period of time the number of patients being treated for sepsis increased from 7% to 15%. The most common secondary diagnoses were PPHN (64%) and air leak syndrome (19%). Table 1 summarizes other features of the patients. ECMO Entry Criteria Criteria for instituting ECMO changed and evolved during the 10 years of this report. The goal of all entry

0

AND

BARTLETT

‘75 ‘76 ‘77 ‘78 ‘79 ‘80 ‘81 ‘82 ‘83 ‘84 ‘85 ‘86 ‘87 ‘88 ‘89

0

Year Fig 3. Annual ECMO patient survival rates and annual total numbers of patients treated through 1989.

criteria was to identify infants with a greater than 80% likelihood of dying. The criteria recognized by the Registry and the incidence in this study were alveolar-arterial oxygen gradient > 600 mm Hg for 6 to 8 hours or more (22%), oxygenation index > 40 for 4 hours (18%), acute deterioration (14%), failure of maximal therapy (34%), barotrauma (l%), cardiac arrest (1%) and others 10%. Arterial blood gas measurements and ventilator settings just prior to ECMO (preductal or postductal not specified) are detailed in Table 1. If one considers the last arterial pre-ECMO blood gas analysis on an annual basis for 1981, 1985, and 1989 (Table 2) there is no significant change in the pH and pC0,. However, the last p0, and its variability may reflect greater consistency and more rigorous application of ECMO therapy. If one considers the last pre-ECMO arterial blood gas analysis by entry diagnosis (Table 2), the following observations can be made. The pOZ for hyaline membrane disease (45 2 23 mmHg) is significantly higher than for other entry 400

20

SNEDECOR,

n MAS n HMD

70

W CDH 300

Ed PPHN

2 3 0 ‘0 t

0

Sepsis

200

f z

0

‘82

‘83

‘84

‘85

‘86

‘87

‘88

‘89

0

1oc

C

73-‘83

‘84

‘85

‘86

‘87

‘88

‘89

Year Fig 2. centers.

ECMO centers opening each year and cumulative

active

Fig 4. sis.

Annual total number of ECMO patients treated, by diagno-

REGISTRY REPORT OF NEONATAL

ECMO

567

Table 1. Patient Characteristics

Table 3. Characteristics of Survivors Versus Nonsurvivors

Male/female (%) Apgar scores

Birthweight (kg)

l-minute

5.2 + 3.8

5-minute

7.2 -t 3.5

Birthweight (kg)

3.2 2 0.6

Gestational age (wk)

39 + 2

Age at ECMO initiation (h)

51.7 t 48.8

ECMO duration (h)

128?

111.6

Last pre-ECMO ABG

3.0 _’ 0.7

<.Ol

30 * 2

38 + 2

<.Ol

Apgar score (5 min)

7+2

622

<.Ol

7.41 + .18

7.29 z .24

pH (last)

40 2 28

PCO, (mm Hg) Outborn (%) ECMO duration (h)

7.39 2 .20

FiO

1.0 f 0.6 rate (breathdmin)

96 t 72

PIP (cm H,O)

45.9 + 10.9

PEEP (cm H,O)

4.3 & 2.8

MAP (cm H,O)

19.7 f 4.8

NOTE. Data given as mean r SD. ABG, arterial

blood gases;

pressure; PEEP, positive end-expiratory

PIP, peak inspiratory

pressure; MAP, mean arterial

pressure.

diagnoses (P < .05); the pC0, (53 + 24 mmHg) for CDH is significantly higher than for other entry diagnoses (P < .Ol); and the pH for sepsis (7.29 + .25) is significantly lower than for other entry diagnoses (P < .OS). ECMO Survival, Clinical Features The overall survival was 83.1% (2,927 survived, 601 died) in a population of infants with predicted mortality of at least 80%. The characteristics of the survivors and nonsurvivors were significantly different (Table 3). ECMO Survival by Entry Criteria Table 4 documents survival as a function of selection criteria. Both a-Ado, gradient and 01 determination yielded an 89% survival. Patients entered because of acute deterioration had a 76% survival. Table 2. Last Pre-ECMO Arterial Blood Gases by Year and Diagnosis PO,

<.Ol

51 ? 32

<.01

83

77

< .05

121 + 60

147 t- 112

< .05

None (85%)

Major (50%)

<.OOl

41 + 24

PCO, (mm Hg)

Abbreviations:

P Value

3.3 * 0.6

Operation

PH Last pre-ECMO ventilator settings Resiiratory

Nonsurvivors

Gestational age (wk)

41 * 32

PO, (mm Hgl

Survivors

lmm

w

~0%

(mmHgi

PH

Year

reflecting the precipitous initiation of ECMO in this group. Although the 34% of patients who entered because of failure of maximal therapy (clinical judgement) might be criticized for lack of rigorously defined entry criteria, this group had only 70% survival, suggesting they were either more desperately ill or ECMO entry was accomplished more, not less, stringently than in the other entry criteria groups. ECMO Survival by EntT Diagnosis Annual survival rates for infants supported with ECMO increased from 56% in 1981 to 86% in 1987, and have remained statistically unchanged since (Fig 3). However, analysis of survival rate by entry diagnosis provides additional detail (Table 5). Analysis of the effect of secondary diagnoses on survival outcome for the primary diagnoses demonstrated two significant associations. Aggregate survival for sepsis as a primary diagnosis decreased from 77% to 63% (P < -05) when the secondary diagnosis was MAS and survival for hyaline membrane disease was decreased from 83% to 75% when the associated secondary diagnosis was sepsis (P < .05). Consideration of patient survival by entry diagnosis and year of entry is more revealing (Fig 5). Patients accrued through 1983 are consolidated as a group because of the small number treated compared with subsequent years. During the last 5 years of this report, patients with MAS and PPHN had a 83% to 94% survival, and survival with sepsis and hyaline membrane disease improved significantIy from 67% and 72%, respectively, to 79% and 88% (P < .05). Current survival rates for MAS, PPHN, hyaline membrane disease,

1981

48 -e 39’

37 2 18

7.31 2 .26

1985

46 t 28”

37 ? 22

7.44 ? .18

1989

37 2 19’

42 lr 22

7.38 + .20

MAS

382

14

37 + 17

7.43 + .I7

HMD

45 2 23t

38 + 14

7.40 + .14

CDH

31 z! 14

53 -c 241

7.32 ? .21

a-Ado,

Sepsis

36 2 21

47 + 33

7.29 2 .21t

01 > 40for4h

PPHN

36 +- 22

40 * 20

7.40 f .21

Acute deterioration

76

Barotrauma

89

Diagnosis

Table 4. Survival by Entry Criteria

NOTE. Data given as mean & SD. Abbreviations: MAS, meconium aspiration syndrome; HMD, hyaline membrane

disease;

CDH, congenital

diaphragmatic

persistent pulmonary hypertension of the newborn.

hernia; PPHN,

Percent > 600 mm Hg for 6 to 8 h

89 89

Maximal therapy failure

70

Cardiac arrest

50

Other

90

STOLAR, SNEDECOR, AND BARTLETT

568

Table 6. Technical Complications

Table 5. Aggregate Survival by Diagnosis MAS

93% (1,262/1,356)

HMD

84% (446/532)

Average Year

PercentWith No Complications

CDH

62% (364/585)

1980

.I2

73

Sepsis

77% (321/416)

1981

.I7

87

PPHN

86% (414/480)

1982

.37

69

Other

77% (120/I 56)

1983

.23

76

1984

.47

65

1985

.28

77

1986

.23

80

1987

.27

77

1988

.27

76

1989

.33

80

.33

73

Abbreviations: MAS, meconium aspiration syndrome; HMD, hyaline membrane

No.

per Patient

disease;

CDH, congenital

diaphragmatic

hernia;

PPHN,

persistent pulmonary hypertension of the newborn.

and sepsis are statistically the same. These are to be distinguished from CDH. The highest survival rate for CDH was 70% in 1987. By 1989, the annual survival rate for CDH had decreased significantly to 56% (P < .Ol). The decrease in annual survival for CDH was also significant when compared with all the other ECMO diagnoses (P < .05).

complications 0.45 +- 0.6.

per patient whereas nonsurvivors had

Technical Complications

Medical Complications

There were 1,039 technical complications in 3,528 ECMO cases, or 0.31 technical complications per infant treated (Table 6). The technical complication rate per patient decreased from .47 per patient (1984) to .33 per patient (1989), whereas the portion of patients without technical complications increased from 65% to 73%. The most common technical complication was incorrect positioning of the cannulae (36%), which then required repositioning. Oxygenator failure represented 24% of the technical complications. Oxygenator failure was managed by either changing oxygenators while on bypass, continuing bypass with compromised oxygenator function, or termination of ECMO support. Although technical complications were not frequent, they did affect survival. Survival with one or more technical complications resulted in an 80% survival compared with 84% survival without complications (P < .Ol). Surviving infants had 0.31 ? 0.5

There were 5,054 medical complications in 3,528 ECMO patients, or 1.50 complications per patient. The medical complication rate per patient improved from 1984 (2.35 per patient) to 1989 (1.61 per patient). Neurological complications (Table 7) predominated (247)o , with a 62% survival (735 of 1,227). Seizures and cerebral infarction documented by either cranial ultrasound or computed axial tomography were the most frequent (78%). Hemorrhagic complications accounted for 21% of medical complications and were evenly distributed among operative site bleeding and hemolysis, with an aggregate survival of 67%. Cardiopulmonary complications accounted for 17% with an aggregate survival of 63%. However, there was a wide survival range from “cardiopulmonary resuscitation required” (45%) to “systemic hypertension” (81%). Renal complications accounted for 14% of the total and had a 60% survival. Almost 50% of the renal complications were managed by either continuous hemofiltration or hemodialysis. Metabolic complication accounted for only 8% of medical complications. Although acidosis and hyperkalemia were associated with a significantly poorer survival than the aggregate population (32% and 50% v 83%), the relative infrequency of persistent metabolic derangements attests to the efficacy of ECMO in correcting derangements and maintaining physiological homeostasis. Overall, 37% of registered ECMO patients had no medical complications reported. In the absence of medical complications, survival was significantly better than when one or more medical complications was present (95% v 76%, P < .Ol). Surviving infants had 1.19 2 1 medical complications and nonsurvivors had 3.11 2 1.

loo-

80.

= L

60.

5 $

-

4o

20-

0’

MAS

-

HMD

-

CDH

-

PPHN

-

Sepsis

r ‘80-‘83

‘84

‘85

‘86

‘87

‘88

Year Fig 5.

Annual survival, by diagnosis, through 1989.

‘89

REGISTRY REPORT OF NEONATAL ECMO

569

Table 7. Complications Type

-

No. of Patients

Neurological 61

6.6

Seizures

526

67.9

Infarction by US

434

51.8

98

82.7

108

65.7

Gastrointestinal

117

63.2

Cannula site

336

67.3

Surgical site

120

55.8

Hemolysis

253

72.3

Other

239

66.9

Brain death

Infarction by CAT Other Hemorrhagic

Cardiopulmonary 8

62.5

CPR required

99

45.5

Myocardial stun

15

80.0

Arrythmia

98

66.3

Symptomatic PDA

23

82.6

Lung hemorrhage

Pulmonary hypertension

124

58.1

Pneumothorax

159

69.8

Other cardiac

343

70.6

Other pulmonary

139

63.3

Creatinine > 1.5 mg%

316

61.7

Dialysis/hemofiltration

344

56.7

73

64.4

Renal

Other Metabolic K’ >6.0

36

50.0

K’ <2.5

82

85.0

Na’ > 150

15

53.0

Na’ ~125

7

57.0

Ca”

9

67.0

Ca’* ~6

43

67.0

CHO > 240

31

48.0

CHO <40

20

45.0

pH >7.60

18

72.0

pH <7.20

34

32.0

Other

96

72.0

>12

Abbreviations:

US, ultrasound; CAT, computed axial tomography;

CPR, cardiopulmonary

resuscitation;

PDA, patent ductus arteriosus,

CHO, carbohydrate.

Diagnosis-Specific Complications

Significant associations between primary ECMO entry diagnoses and specific complications were limited to CDH and sepsis. CDH, compared with MAS, PPHN, and sepsis, was significantly more often associated with hemorrhage (operative sites, gastrointestinal, other) and pulmonary vasodilator use (P < .Ol). Sepsis, compared with MAS, PPHN, and CDH, was significantly more frequently associated with seizures, renal failure, culture-positive infection, neutropenia, intracranial hemorrhage, and brain death (P < .Ol). DISCUSSION Between

registered

1980 and 1989, 3,528 ECMO cases were with an overall survival of 83% in a

population expected to have at least an 80% mortality. Although previous Registry reports3 described a learning curve for the first 10 patients, with subsequent experience and additional new ECMO centers the overall survival rate has been unaffected and remains close to 90% for all entry diagnoses except CDH. In addition to the sustained survival rate, the annual rate of technical complications has decreased over this time from .47 to .33 per patient, indicating an increased number of patients without technical complications. Similarly, the annual rate of medical complications decreased from 2.35 to 1.60 per patient. Consequently, despite any learning curve effects that might accompany new ECMO centers and rapidly expanding numbers of infants treated, it is encouraging to see decreasing morbidity and sustained excellent survival. Diagnosis-specific outcome analysis suggests more about the particular diagnosis than ECMO, per se. Specifically, CDH had significantly more hemorrhagic complications and poorer survival than other diagnoses. Of all the ECMO entry diagnoses, CDH is the only situation in which the patient has less than two completely developed lungs and requires an operation. The hemorrhagic complications can be attributed to the major operation necessary to repair the diaphragmatic hernia, usually within 96 hours of ECMO initiation. The poorer survival can be related, in part, to the hemorrhagic complications, but also to pulmonary hypoplasia to a degree that is incompatible with life. Initially, ECMO was used for infants with CDH primarily if they had shown some evidence of adequate lung parenchyma before developing refractory pulmonary hypertension. In the last few years some centers have used ECMO for CDH infants and intractable pulmonary hypertension without consideration of the lung parenchyma. This may have resulted in more CDH/ECMO deaths from irreverable pulmonary hypoplasia. Because more ECMO centers try to determine the degree of pulmonary hypoplasia prior to ECMO, the mortality may decrease. Further, with the increased availability of ECMO, more infants with CDH, particularly the most desperately ill, will be subjected to ECMO. Although survival from sepsis is not as obviously different from the aggregate as is survival from CDH, the morbidity from sepsis does differ. For all other ECMO entry diagnoses, the lungs and, perhaps, the heart and kidneys, are the target organs. Sepsis is much more global in its effects and is characterized by multiple organ failure, which accounts for the increased morbidity.* Selection criteria remain problematic for a variety of reasons. They cannot be viewed as absolute because of variability between centers. What represents

570

STOLAR, SNEDECOR, AND BARTLETT

likely 80% mortality in one center may not apply to another. Historical controls are misleading because changing respiratory therapy strategies make historical populations difficult to compare. Also, once an ECMO center becomes established, a more challenging group of patients will be attracted than previously was the case. Further, as consideration of the last pre-ECMO arterial blood gas suggested, a single entry criteria cannot be generalized for all entry diagnoses. It is misleading to suggest that criteria for an 80% predicted mortality is the same for MAS, CDH, PPHN, and sepsis.

APPENDIX:

Subsequent patients registered in the Neonatal ECMO registry of the Extracorporeal Life Support Organization will address these issues more thoroughly, as specific details of the pre-ECMO condition and therapeutic strategies are collected. The multicenter data yield information not available from a single institutional experience. Although entry criteria and conventional therapy continue to evolve, ECMO currently improves survival from an estimated 20% to 83% overall. Individual prognosis depends on the specific diagnosis, clinical condition of the patient, and presence or absence of complications.

CONTRIBUTING

Children’s Hospital of Alabama, Birmingham, AL St Vincent’s Hospital, Birmingham, AL Phoenix Children’s Hospital, Pheonix, AZ St Joseph’s Hospital and Medical Center, Pheonix, AZ Arkansas Children’s Hospital, Little Rock, AR Children’s Hospital of Los Angeles, Los Angeles, CA Children’s Hospital of Oakland, Oakland, CA Children’s Hospital of Orange County, Orange, CA Huntington Memorial Hospital, Pasadena, CA San Diego Regional ECMO Program, San Diego, CA Sutter Memorial Hospital, Sacramento, CA Stanford University Hospital, Stanford, CA UCLA Medical Center, Los Angeles, CA University of California San Francisco, San Francisco, C‘A Children’s Hospital, Denver, CO Children’s National Medical Center, Washington, DC Georgetown University Hospital, Washington, DC Miami Children’s Hospital, Miami, FL University of Florida, Gainsville, FL Arnold Palmer Hospital, Orlando, FL Medical College of Georgia, Augusta, GA St Luke’s Hospital, Boise, ID Children’s Memorial Hospital, Chicago, IL Cook County Hospital, Chicago, IL Lutheran General Hospital, Park Ridge, IL St Francis Medical Center, Peoria, IL James Whitcomb Riley Hospital. Indianapolis, IN Kosair Children’s Hospital, Louisville, KY Oschner Foundation Hospital, New Orleans, LA Boston Children’s Hospital, Boston, MA Massachusetts General Hospital, Boston, MA Johns Hopkins University, Baltimore, MD Children’s Hospital of Michigan, Detroit, MI

ECMO CENTERS University of Michigan Medical Center, Ann Arbor, MI Minnesota Regional ECMO Program, Minneapolis, MN Cardinal Glennon Children’s Hospital, St Louis, MO The Children’s Mercy Hospital, Kansas City, MO St Louis Children’s Hospital, St Louis, MO University of Nebraska Medical Center, Omaha, NE The Babies Hospital, New York, NY Charlotte Memorial Hospital, Charlotte, NC Children’s Hospital Medical Center, Cincinatti, OH Miami Valley Hospital, Dayton, OH Rainbow Babies and Childrens Hospital, Cleveland, OH Children’s Hospital, Columbus, OH Eastern Oklahoma Perinatal Center, Tulsa, OK Emanuel Hospital and Health Center, Portland, OR Children’s Hospital of Pittsburgh, Pittsburgh, PA Thomas Jefferson University Hospital, Philadelphia, PA Medical University of South Carolina, Charleston, SC University of South Carolina, Columbia, SC John Sealy Hospital, Galveston, TX Texas Tech University, Lubbock, TX Wilford Hall Medical Center, Lackland Air Force Base, TX Presbyterian Hospital of Dalklas, Dallas, TX Primary Children’s Medical Center, Salt Lake City, UT Medical College of Virginia, Richmond, VA Children’s Hospital and Medical Center, Seattle, WA Children’s Hospital of Wisconsin, Milwaukee, WI Royal Alexandra Children’s Pavillion, Edmonton, Alberta Assistance Publique Hospitauz de Paris, Paris, France Hospital Trousseau, Paris, France Universitats-Kinderklinik, Mannheim, Germany Kumamoto University Medical School, Kumamoto, Japan Central Hospital, Kasugai’aichi, Japan Karolinska Institutet, Stockholm, Sweden

REFERENCES 1. Bartlett RH, Gazanniga AB, Toomasian JT, et al: Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure. Ann Surg 204:236-243, 1986 2. Loe WA, Graves ED, Oschner JL, et al: Extracorporeal membrane oxygenation for newborn respiratory failure. J Pediatr Surg 20:684-690, 1985 3. Krummel TM, Greenfield LJ, Kirkpatrik BU, et al: Extracor-

poreal membrane oxygenation Pediatr Ann 11:905-909, 1982 4. White JJ, Andrews HG, respiratory support in newborn ator. Surgery 70:288-292,197l 5. Toomasian JT, Snedecor experience with extracorporeal

in neonatal

pulmonary failure.

Zisemberg H, et al: Prolonged infants with a membrane oxygenSM, Cornell RG, et al: National membrane oxygenation for new-

REGISTRY REPORT OF NEONATAL

ECMO

born respiratory failure. Tram Am Sot Artif Int Org 34:140-145, 1988 6. Bartlett RH, Roloff DW, Cornell RG, et al: Extracorporeal membrane oxygenation in neonatal respiratory failure: A prospective randomized study. Pediatrics 76:479-482, 1984 7. O’Rourke PP, Crone RK, Vacante JP, et al: Extracorporeal membrane oxygenation and conventional medical treatment: A prospective randomized study. Pediatrics 84:957, 1989 8. Bartlett RH, Gazanniga AB, Huxtable RF, et al: Extracorpo-

571

real circulation (ECMO) in neonatal respiratory failure. J Thorac Cardiovasc Surg 76:826,1977 9. Chapman R, Toomasian JT, Bartlett, RH: ECMO Technical Specialist Manual. Ann Arbor, MI, University of Michigan Press, 1989 lo.-Schmelling D, Chapman R, Bartlett RH. et al: ECMO support for neonatal sepsis. Presented at the Fourth Annual Children’s Hospital National Medical Center ECMO Conference, Snowmass, CO (abstr)