Aust. Paediatr. J .
(1979), 15:233-237
Iatrogenic Hazards of Neonatal Intensive Care in Extremely Low Birthweight Infants V.Y.H. YU1, P.H. HEWSON*, and E. HOLLINGSWORTH3. Department of Paediatrics, Queen Victoria Medical Centre, Melbourne, Victoria.
Yu, V.Y.H., Hewson, P.H. and Hollingsworth, E. (1979). Aust. Paediatr. J., 15, 233-237. Iatrogenic hazards of neonatal intensive care in extremely low birthweight infants. The history of neonatal paediatrics has evidenced over-enthusiasm in the introduction of innovations in management which are almost always accompanied by unforseen hazards. We have scrutinised present practices in the intensive care of 55 infants weighing < 1000 g at birth, which might be responsible for the current trends in iatrogenic disease patterns. Umbilical arterial catheterization was associated with a 29% complication rate. Hyperoxia was documented in 81% of infants during oxygen therapy. Retrolental fibroplasia was present in 1 survivor. Blood sampling for microbiochemical and haematological monitoring increased the need for replacement blood transfusions. Postnatally acquired cytomegalovirus infection occurred in 4 infants, 3 of whom were probably infected from transfused blood. Complications associated with current nutritional management included hyperglycaemia, hypocalcaemia and hyponatraemia. Iatrogenic factors might have contributed to the occurrence of necrotizing enterocolitis, patent ductus arteriosus and cholestatic jaundice. We conclude that current neonatal intensive care practices need ongoing scrutiny and great caution should be taken in introducing new advances.
Mortality and morbidity rates in low birthweight infants have altered dramatically pver the last decade. While modern obstetrics and neonatal intensive care have an important contribution to the increased survival rate and a decline in physical and mental handicaps, iatrogenic hazards as a result of over-enthusiasm in the introduction of new practices continue to be documented. Past hazards have included the dangers of retrolental fibroplasia (RLF) from indiscriminate' use of oxygen therapy followed by the period of irrational restriction of oxygen and its consequences (Cross, 1973); of starvation dehydration and hypoglycaemia from the practice of delayed onset of feeding . (Smallpiece and Davies, 1964); of induced hypothermia (Silverman et al, 1958); of water-soluble vitamin K producing kernicterus (Meyer and Angus, 1956); and of chloramphenicol intoxication leading to the gray syndrome (Sutherland, 1959). Although important advances have been made in the understanding of the pathophysiology in low birthweight infants which have led to innovations in management, iatrogenic hazards continue to threaten certain modern aspects of neonatal intensive care. The purpose of this report is to document the extent of 'exposure to possible hazards which resulted from current established practices in a group of extremely low birthweight (ELBW) infants, as part of an ongoing clinical audit essential to every neonatal intensive care programme. 1. Director ofNeonatal Intensive Care. 2. Paediatric Registrar. 3. Charge Sister, Neonatal Intensive Care Unit
First received April 30, 1979.
PATIENTS AND METHODS The patients reviewed came from the population of 55 ELBW infants weighing < lOOOg at birth admitted to the Queen Victoria Medical Centre during 1977 and 1978. The composition of the population and its general management was described in a previous report (Yu and Hollingsworth, 1979a). In summary, the mean birthweight was 830g (range 510-1000 g) and the mean gestational age was 26 weeks (range 23-28 weeks). Overall neonatal survival rate was 60%; 44% for the 16 infants in the 501-75Og weight group and 67% for the 39 infants in the 751-1OOOg weight group. Neonatal practices which might be responsible for possible iatrogenic hazards during the two study years were as follows. Umbilical arterial catheters were Inserted for the purpose of blood sampling, blood pressure monitoring and parenteral infusidn, the latter consisting of glucose or glucose-aminoacid and fat solutions containing 1 unit per ml of heparin. Alkali therapy was used if the pH falls below 7.25 in association with a base deficit of more than 6 mmolll. A bolus of hypertonic sodium bicarbonate was never administered except in an emergency such as cardiac arrest or extreme bradycardia. If base was required, a slow infusion over 2 to 3 hours of sodium bicarbonate at a concentration of 2.8% or less was given. The protocol for oxygen, and ventilation therapy has been detailed by Yu and Hollingsworth (1979b). Blood gas monitoring (25 pl blood) was carried out every 3-6 hours and following changes in oxygen or ventilatory therapy. Weekly examinations of the fundi were carried out by indirect opthalmoscopy for changes of retrolental fibroplasla. Biochemical and haematological monitoring (100 pl blood)
234
AUSTRALIAN PAEDlATRlC JOURNAL
TABLE I Postnatal cytomegalovirus infection. ~~
~~
NO.
~
~
(g/wk)
1 2 3 4
~~
BIRTHWEIGHT AGE CMV HEPATONO. OF VOLUME OF GESTATION ISOLATED SPLENOBLOOD BLOOD IN URINE MEGALY TRANSFUSION TRANSFUSION 880127 790127 810125 700127
(wk)
(wk)
5 6
6 6
10 10
10 -
which included serum osmolality, glucose, sodium, potassium, calcium, bilirubin (total and direct), haemoglobin and haematocrit was usually carried out daily in the first week and at less frequent intervals subsequently. Blood transfusions, by slow infusions of 10-15 m l l k g of fresh blood or packed cells, were given to ELBW infants with respiratory distress or preterm recurrent apnoea when the haematocrit was less than 40%, when the systolic blood pressure was less than 40 mmHg and when greater than 10% of the blood volume was lost from blood sampling. Infants had fortnightly urine and saliva specimens taken for cytomegalovirus (CMV) culture. Prophylactic antibiotics were not given for prolonged rupture of membranes or endotracheal intubation and ventilation, but only on clinical suspicion of sepsis. Phototherapy was commenced at a serum indirect bilirubin level of 150 pmolll or earlier if excessive bruising or haemolytic disease was present. Exchange transfusion was carried out at a serum indirect billirubin of 200-300 pmolll depending on co-existing factors influencing the risk of kernicterus. Infants were fed fresh breast milk from their own mothers when possible. Parenteral nutrition was used prior to establishment of enteric feeding. Details of its formulation and protocol for administration were previously reported (Yu al, 1979a). The total fluid administered to ELBW infants in the first week was 100-150 mllkgl24h and the volume was increased to 150-200 mllkgl24h during the second week. This varied in the individual infant according to clinical and biochemical hydration state. Parents, siblings and the extended family were allowed unrestricted contact with their infants (Yu, 1977).
RESULTS Oxygen and ventilation therapy. All but 2 infants received oxygen therapy. Hyperoxaemia (Pa02 > 100mmHg) was documented on one or more occasions in 43 (81%) infants. RLF was diagnosed in one survivor with neovascularization and peripheral clouding in one eye and peripheral retinal detachment in the other, equivalent to Kingham (1977) Stages II and 111 changes respectively. Seventeen of the 51 infants who required assisted ventilation developed pulmonary interstitial emphysema, a condition which was associated with an increased incidence of pneumothorax (4 infants) and bronchopulmonary dysplasia (Northway et a/, 1967, Stage 111 in 8 infants and Stage IV in 2 infants). Three infants with Stage 111 changes died. Survivors with bronchopulmonary dysplasia required up to 76 days of assisted ventilation and 84 days of oxygen therapy. The morbidity from oxygen and ventilation has been described in detail in a previous paper (Yu and Hollingsworth, 1979b).
(ml)
4
5 6 1
63 79 93 9
Alkali therapy. All infants, with the exception of 1 who died within 30 minutes after birth had alkali therapy. During their nursery stay, the mean amount of sodium bicarbonate given was 13 mmol (range 1-45 mmol). Fourteen infants (25%) had documented hypernatraemia (serum sodium > 150 rnmolll) had hyperosmolality (serum osmolality and 16 infants (19Y0) > 320 mOsmlkg H,O). Eight of the 14 infants (57%) with hypernatraemia had intraventricular haemorrhage compared with 8 of the 41 infants (22%) without hypernatraemia who developed intraventricular haemorrhage. The difference is statistically significant (X2 = 5.5, p < 0.02). The incidence of intraventricular haemorrhage in infants with or without hyperosmolality was both 29%. Umbilical arterial catheterization. This was attempted in all infants except for 2 who died within 2 hours of birth. Five of the 53 (9%) catheterization attempts failed, for which a umbilical venous catheter was inserted with placement of its tip in the right atrium. Arterial catheters, inserted'in the abdominal aorta to the level of third ,or fourth lumbar vertebrae, were left in situ for between 5 to 10 days. Nine infants had transient ischaemia of the toes which resolved with removal of the catheter. Chlorhexidine 0.5% and Alcohol 70%, the antiseptic used in conjunction with the catheterization procedure, produced skin burns in 2 infants. Two infants had a haemorrhage estimated at about 5 mls due to accidental disconnection of the catheter from the pressure transducer line. Mural thrombus in the abdominal aorta was found at necropsy in 2 infants. Though none of the complications had contributed to the mortality in the series, the observed morbidity rate was 29%. Blood loss and replacement transfusions. Repeated blood sampling in 51 infants amounted to a mean of 38 mls (range 2 to 91 mls) even with microanalysis techniques. This has in part contributed to the increased need for 'top-up' blood transfusions. Forty-eight (87%) of the infants had 1 to 12 transfusions (mean 4) which gave a cumulative volume of 9 to 163 mls (mean 59 mls). The majority of the donor blood was transfused within 24 hours of collection and CMV screening was not performed on the donor's blood. Four infants had CMV isolated in the urine (Table I). Three of these infants were probably infected from their multiple blood transfusions. Except for transient hepatosplenomegaly in 3 infants, they were asymptomatic to date. One infant who had CMV isolated in the urine for the first time at 10 weeks of age also had excessive increase in head circumference from 5 weeks of age, which investigations showed was due to porencephaly. Although no neurological abnormality was obvious at 6 months of age, long term prognosis is still to be determined. Nutrition and fluid intake. The mode of oral feeding, number of infants and period on nil-orally and parenteral nutrition
235
IATROGENIC HAZARDS OF NEONATAL INTENSIVE CARE
T A B L E II Nutritional intake data.
NEONATAL SURVIVORS
NEONATAL DEATHS
TOTAL
(n=33)
(n=22)
(n=55)
MODE OF O R A L FEEDING Expressed breast m i l k Formula Nil -orally PERIOD ON NO O R A L FEEDS No. of infants No. of days (mean%em) (range) ~~
23 10 0 29 12 5 2.0 1-41
22 4 '0.8 0.02 - 14
51 (93%) 951.4 0.02 - 41
29 20 k 2.6 2 - 50
12 8 k2.2 1-22
41 (75%) 17 5 2.1 1 - 50
are summarised in Table 11. Three infants (5%) had hypoglycaemia (blood glucose < 1.5 m m o l / l ) but 19 infants (35%) had hyperglycaemia (blood glucose > 1 1 mmolll). Although intravenous calcium was provided routinely at a rate of 1 m m o l / k g / 2 4 hrs., 22 infants (40%) had hypocalcaemia (serum calcium
(1979), 15:233-237
Iatrogenic Hazards of Neonatal Intensive Care in Extremely Low Birthweight Infants V.Y.H. YU1, P.H. HEWSON*, and E. HOLLINGSWORTH3. Department of Paediatrics, Queen Victoria Medical Centre, Melbourne, Victoria.
Yu, V.Y.H., Hewson, P.H. and Hollingsworth, E. (1979). Aust. Paediatr. J., 15, 233-237. Iatrogenic hazards of neonatal intensive care in extremely low birthweight infants. The history of neonatal paediatrics has evidenced over-enthusiasm in the introduction of innovations in management which are almost always accompanied by unforseen hazards. We have scrutinised present practices in the intensive care of 55 infants weighing < 1000 g at birth, which might be responsible for the current trends in iatrogenic disease patterns. Umbilical arterial catheterization was associated with a 29% complication rate. Hyperoxia was documented in 81% of infants during oxygen therapy. Retrolental fibroplasia was present in 1 survivor. Blood sampling for microbiochemical and haematological monitoring increased the need for replacement blood transfusions. Postnatally acquired cytomegalovirus infection occurred in 4 infants, 3 of whom were probably infected from transfused blood. Complications associated with current nutritional management included hyperglycaemia, hypocalcaemia and hyponatraemia. Iatrogenic factors might have contributed to the occurrence of necrotizing enterocolitis, patent ductus arteriosus and cholestatic jaundice. We conclude that current neonatal intensive care practices need ongoing scrutiny and great caution should be taken in introducing new advances.
Mortality and morbidity rates in low birthweight infants have altered dramatically pver the last decade. While modern obstetrics and neonatal intensive care have an important contribution to the increased survival rate and a decline in physical and mental handicaps, iatrogenic hazards as a result of over-enthusiasm in the introduction of new practices continue to be documented. Past hazards have included the dangers of retrolental fibroplasia (RLF) from indiscriminate' use of oxygen therapy followed by the period of irrational restriction of oxygen and its consequences (Cross, 1973); of starvation dehydration and hypoglycaemia from the practice of delayed onset of feeding . (Smallpiece and Davies, 1964); of induced hypothermia (Silverman et al, 1958); of water-soluble vitamin K producing kernicterus (Meyer and Angus, 1956); and of chloramphenicol intoxication leading to the gray syndrome (Sutherland, 1959). Although important advances have been made in the understanding of the pathophysiology in low birthweight infants which have led to innovations in management, iatrogenic hazards continue to threaten certain modern aspects of neonatal intensive care. The purpose of this report is to document the extent of 'exposure to possible hazards which resulted from current established practices in a group of extremely low birthweight (ELBW) infants, as part of an ongoing clinical audit essential to every neonatal intensive care programme. 1. Director ofNeonatal Intensive Care. 2. Paediatric Registrar. 3. Charge Sister, Neonatal Intensive Care Unit
First received April 30, 1979.
PATIENTS AND METHODS The patients reviewed came from the population of 55 ELBW infants weighing < lOOOg at birth admitted to the Queen Victoria Medical Centre during 1977 and 1978. The composition of the population and its general management was described in a previous report (Yu and Hollingsworth, 1979a). In summary, the mean birthweight was 830g (range 510-1000 g) and the mean gestational age was 26 weeks (range 23-28 weeks). Overall neonatal survival rate was 60%; 44% for the 16 infants in the 501-75Og weight group and 67% for the 39 infants in the 751-1OOOg weight group. Neonatal practices which might be responsible for possible iatrogenic hazards during the two study years were as follows. Umbilical arterial catheters were Inserted for the purpose of blood sampling, blood pressure monitoring and parenteral infusidn, the latter consisting of glucose or glucose-aminoacid and fat solutions containing 1 unit per ml of heparin. Alkali therapy was used if the pH falls below 7.25 in association with a base deficit of more than 6 mmolll. A bolus of hypertonic sodium bicarbonate was never administered except in an emergency such as cardiac arrest or extreme bradycardia. If base was required, a slow infusion over 2 to 3 hours of sodium bicarbonate at a concentration of 2.8% or less was given. The protocol for oxygen, and ventilation therapy has been detailed by Yu and Hollingsworth (1979b). Blood gas monitoring (25 pl blood) was carried out every 3-6 hours and following changes in oxygen or ventilatory therapy. Weekly examinations of the fundi were carried out by indirect opthalmoscopy for changes of retrolental fibroplasla. Biochemical and haematological monitoring (100 pl blood)
234
AUSTRALIAN PAEDlATRlC JOURNAL
TABLE I Postnatal cytomegalovirus infection. ~~
~~
NO.
~
~
(g/wk)
1 2 3 4
~~
BIRTHWEIGHT AGE CMV HEPATONO. OF VOLUME OF GESTATION ISOLATED SPLENOBLOOD BLOOD IN URINE MEGALY TRANSFUSION TRANSFUSION 880127 790127 810125 700127
(wk)
(wk)
5 6
6 6
10 10
10 -
which included serum osmolality, glucose, sodium, potassium, calcium, bilirubin (total and direct), haemoglobin and haematocrit was usually carried out daily in the first week and at less frequent intervals subsequently. Blood transfusions, by slow infusions of 10-15 m l l k g of fresh blood or packed cells, were given to ELBW infants with respiratory distress or preterm recurrent apnoea when the haematocrit was less than 40%, when the systolic blood pressure was less than 40 mmHg and when greater than 10% of the blood volume was lost from blood sampling. Infants had fortnightly urine and saliva specimens taken for cytomegalovirus (CMV) culture. Prophylactic antibiotics were not given for prolonged rupture of membranes or endotracheal intubation and ventilation, but only on clinical suspicion of sepsis. Phototherapy was commenced at a serum indirect bilirubin level of 150 pmolll or earlier if excessive bruising or haemolytic disease was present. Exchange transfusion was carried out at a serum indirect billirubin of 200-300 pmolll depending on co-existing factors influencing the risk of kernicterus. Infants were fed fresh breast milk from their own mothers when possible. Parenteral nutrition was used prior to establishment of enteric feeding. Details of its formulation and protocol for administration were previously reported (Yu al, 1979a). The total fluid administered to ELBW infants in the first week was 100-150 mllkgl24h and the volume was increased to 150-200 mllkgl24h during the second week. This varied in the individual infant according to clinical and biochemical hydration state. Parents, siblings and the extended family were allowed unrestricted contact with their infants (Yu, 1977).
RESULTS Oxygen and ventilation therapy. All but 2 infants received oxygen therapy. Hyperoxaemia (Pa02 > 100mmHg) was documented on one or more occasions in 43 (81%) infants. RLF was diagnosed in one survivor with neovascularization and peripheral clouding in one eye and peripheral retinal detachment in the other, equivalent to Kingham (1977) Stages II and 111 changes respectively. Seventeen of the 51 infants who required assisted ventilation developed pulmonary interstitial emphysema, a condition which was associated with an increased incidence of pneumothorax (4 infants) and bronchopulmonary dysplasia (Northway et a/, 1967, Stage 111 in 8 infants and Stage IV in 2 infants). Three infants with Stage 111 changes died. Survivors with bronchopulmonary dysplasia required up to 76 days of assisted ventilation and 84 days of oxygen therapy. The morbidity from oxygen and ventilation has been described in detail in a previous paper (Yu and Hollingsworth, 1979b).
(ml)
4
5 6 1
63 79 93 9
Alkali therapy. All infants, with the exception of 1 who died within 30 minutes after birth had alkali therapy. During their nursery stay, the mean amount of sodium bicarbonate given was 13 mmol (range 1-45 mmol). Fourteen infants (25%) had documented hypernatraemia (serum sodium > 150 rnmolll) had hyperosmolality (serum osmolality and 16 infants (19Y0) > 320 mOsmlkg H,O). Eight of the 14 infants (57%) with hypernatraemia had intraventricular haemorrhage compared with 8 of the 41 infants (22%) without hypernatraemia who developed intraventricular haemorrhage. The difference is statistically significant (X2 = 5.5, p < 0.02). The incidence of intraventricular haemorrhage in infants with or without hyperosmolality was both 29%. Umbilical arterial catheterization. This was attempted in all infants except for 2 who died within 2 hours of birth. Five of the 53 (9%) catheterization attempts failed, for which a umbilical venous catheter was inserted with placement of its tip in the right atrium. Arterial catheters, inserted'in the abdominal aorta to the level of third ,or fourth lumbar vertebrae, were left in situ for between 5 to 10 days. Nine infants had transient ischaemia of the toes which resolved with removal of the catheter. Chlorhexidine 0.5% and Alcohol 70%, the antiseptic used in conjunction with the catheterization procedure, produced skin burns in 2 infants. Two infants had a haemorrhage estimated at about 5 mls due to accidental disconnection of the catheter from the pressure transducer line. Mural thrombus in the abdominal aorta was found at necropsy in 2 infants. Though none of the complications had contributed to the mortality in the series, the observed morbidity rate was 29%. Blood loss and replacement transfusions. Repeated blood sampling in 51 infants amounted to a mean of 38 mls (range 2 to 91 mls) even with microanalysis techniques. This has in part contributed to the increased need for 'top-up' blood transfusions. Forty-eight (87%) of the infants had 1 to 12 transfusions (mean 4) which gave a cumulative volume of 9 to 163 mls (mean 59 mls). The majority of the donor blood was transfused within 24 hours of collection and CMV screening was not performed on the donor's blood. Four infants had CMV isolated in the urine (Table I). Three of these infants were probably infected from their multiple blood transfusions. Except for transient hepatosplenomegaly in 3 infants, they were asymptomatic to date. One infant who had CMV isolated in the urine for the first time at 10 weeks of age also had excessive increase in head circumference from 5 weeks of age, which investigations showed was due to porencephaly. Although no neurological abnormality was obvious at 6 months of age, long term prognosis is still to be determined. Nutrition and fluid intake. The mode of oral feeding, number of infants and period on nil-orally and parenteral nutrition
235
IATROGENIC HAZARDS OF NEONATAL INTENSIVE CARE
T A B L E II Nutritional intake data.
NEONATAL SURVIVORS
NEONATAL DEATHS
TOTAL
(n=33)
(n=22)
(n=55)
MODE OF O R A L FEEDING Expressed breast m i l k Formula Nil -orally PERIOD ON NO O R A L FEEDS No. of infants No. of days (mean%em) (range) ~~
23 10 0 29 12 5 2.0 1-41
22 4 '0.8 0.02 - 14
51 (93%) 951.4 0.02 - 41
29 20 k 2.6 2 - 50
12 8 k2.2 1-22
41 (75%) 17 5 2.1 1 - 50
are summarised in Table 11. Three infants (5%) had hypoglycaemia (blood glucose < 1.5 m m o l / l ) but 19 infants (35%) had hyperglycaemia (blood glucose > 1 1 mmolll). Although intravenous calcium was provided routinely at a rate of 1 m m o l / k g / 2 4 hrs., 22 infants (40%) had hypocalcaemia (serum calcium
