| Why Mothers Die | ||||
|
CHAPTER 17 INTENSIVE CARE There were 107 maternal deaths, including seven Fortuitous and six Late, in which there was a recorded need for intensive care in this triennium. These are shown in Table 17.1. Due to the incomplete forms for some other cases it is likely that the total number is actually higher.
Intensive care services were required for a range of reasons, from resuscitation for a few hours only to over 40 days of treatment for multiple organ failure, often following acute respiratory distress syndrome (ARDS).
There is a discrepancy between the types of condition causing mortality and those causing ICU admission. This is because fatal conditions may cause rapid death before admission to the ICU and because Late maternal deaths are often not reported to this Enquiry. In addition women with some specific disorders are cared for on specialist units, and those with medical disorders are often managed on medical wards.
Direct causes of death
Among the Direct deaths, 75% (9) of the women who suffered a haemorrhage required intensive care and 50% (7) of those with sepsis; 35% (7) of women with pregnancy-induced hypertension and 35% ( 6) of those with amniotic fluid embolism survived long enough to be transferred to an ICU. Not surprisingly, given the speed of collapse following pulmonary embolism, only 19% (9) of these women were admitted to an ICU. Half of the women who died in early pregnancy also required such support. Many of these women died from ARDS, which is still associated with an adverse outcome.
Indirect causes of death
Of the 42 women who had an Indirect death on an ICU, the largest number were due to "other Indirect" causes of maternal death, as discussed in Chapter 11. These cover a wide range of medical and surgical conditions. Fortuitous and Late deaths
Seven deaths on ICU were Fortuitous, and six Late. The causes of the Late deaths included cystic fibrosis, cerebral haemorrhage and neoplastic disease. Some Indirect, Late and Fortuitous deaths associated with ICU are difficult to identify, as in many cases the medical attendants may have forgotten that the patient had been pregnant within the time defined for Late deaths, or are not aware of the need to report a maternal death.
Table 17.1 - Number Of Direct and Indirect cases admitted to ICU by Chapter; United Kingdom 1994-96.
General discussion
The assessment of these cases and a review of the published literature (see the Annex to this Chapter) allow the following conclusions to be drawn.
The development of multiple organ failure is often preceded by ARDs. Pregnant women are at risk of developing ARDS from obstetric complications such as AFE, sepsis, pre-eclampsia, hypertensive disorders of pregnancy, placental abruption and dead fetus syndrome, as well as being at increased risk from some non-obstetric conditions. Pregnancy predisposes to other pulmonary insults, such as aspiration of stomach contents, pneumonia, air embolism, massive haemorrhage and non-obstetric infections.
The differential diagnosis of ARDS in pregnancy includes venous thromboembolism, AFE, pulmonary oedema secondary to pre-eclampsia, tocolytic pulmonary oedema, aspiration pneumonitis, peripartum cardiomyopathy, pneumomediastinum, air embolism, asthma, pneumonia and cardiac disease. Asthma is associated with prematurity, low birth weight and increased perinatal mortality, probably due to poor asthma control but pregnancy is not a contraindication to steroid treatment.
Tocolytic-induced pulmonary oedema is due to administration of beta-adrenergic agents especially terbutaline and ritodrine which are used to inhibit uterine contractions and their use in pregnancies may be associated with hypokalaemia, hyperglycaemia, tachyarrythmia and sodium retention as well as pulmonary oedema.
Pregnancy predisposes to infection and its increased severity. These include pneumonias due to viruses such as varicella and herpes simplex.
The consensus remains that provision of good intensive and high-dependency care reduces mortality and that the best chance of success occurs when patients are treated early in their illness.
ANNEX TO CHAPTER 17 RECENT REVIEWS OF MATERNAL MORTALITY AND MORBIDITY ASSOCIATED WITH ICU Since the last triennial Report several centres have reviewed the requirement of pregnant patients for intensive care.
Lapinsky et al. reviewed 65 obstetric admissions to an ICU in an academic hospital over a five-year period (0.26% of deliveries)1. None of these patients died. Umo-Etuk et al. undertook a five-year review and found 39 parturient patients were admitted to their general ICU in a five-year period2. The authors suggest from their review that the fall in maternal mortality reflects an improvement in organ support in the ICU but as a result there is an increase in the number of deaths from ARDS. It is recognised that it is much more difficult to measure morbidity (and set standards of care) but admission to the ICU identifies a subset of parturient women at risk of severe morbidity. In this study the majority of patients were admitted to the ICU either as a direct result of pregnancy or because of a medical or surgical problem which was aggravated by the physiological problems of pregnancy.
Bouvier-Colle et al. studied 435 obstetric patients admitted to ICU3 and calculated that the frequency was 36 per 100,000 live births. The mortality was lower with scheduled maternity cases in a teaching hospital and these authors concluded that most obstetric patients with serious diseases were referred for suitable care.
Wheatley et al. reviewed admissions to their ICU to see whether admission could have been predicted4. They found that 67% of patients had no previous medical or obstetric history. As in other series, the major reasons for admission were hypertensive disorders of pregnancy (66%) and haemorrhage (19%); 79% followed caesarean section and 40% required ventilatory support. The perinatal mortality was 6% and there were three maternal deaths. The need for ICU admission was unpredictable in two-thirds of cases. The authors suggest that a small proportion of women who develop complications of pregnancy (0.1-0.9%) require admission to an ICU. Homerton hospital, from which this survey comes, is in a deprived inner city area. The high rate of ICU obstetric admissions (0.75%) is similar to other inner city areas but for Nottingham during the period 1982-86 it was only 0.1%. A definitive diagnosis of ARDS was only made in two patients both of whom died after amniotic fluid embolism (AFE); a low incidence compared to other published series. There has been a recent trend to increase use of regional techniques for hypertensive disease but, as yet no significant demonstrable difference in fetal or maternal outcome.
A policy of early intervention and treatment on a multidisciplinary basis, which may involve intubation, ventilation, invasive monitoring and vasoactive drugs, was used preventively as well as after the onset of problems. This approach to management by early involvement of all relevant specialties to provide optimal care can alleviate the progression to multiple organ failure, and improve prognosis. They admit that high-dependency unit (HDU) care may have been just as effective for patients who are conscious and have single organ dysfunction.
Stephens reviewed hospital records of obstetric patients from a nearby hospital admitted to ICU for respiratory support after an anaesthetic complication5. In a 10-year period there were 126 obstetric admission to the ICU from 61,435 deliveries, of which 16 were due to anaesthetic complications, 12 after general and four after regional anaesthesia. Complications included anaphylaxis, high spinal block and failure of endotracheal intubation. The incidence of major complication causing admission to ICU was 1: 932 after general anaesthesia and for regional anaesthesia 1: 4177 when these were given for delivery. If a complication requiring ICU admission and mechanical ventilation is used as the criterion of safety it appears that regional anaesthesia is safer than general anaesthesia for delivery.
Severe maternal morbidity is easy to underestimate because pregnant women are usually healthy and recover quickly, and are discharged with relatively little follow-up. A review by Bewley & Creighton identified a small but very sick group of pregnant women with high rates of medical intervention, many of whom did not go home with live babies or with their fertility intact6. They required a disproportionate amount of resources and skill and are women for whom medical facilities may have been life-saving. The definition of a "near miss" may be identified by severity or by disease - in this case, admission to ICU was chosen as it was easy to measure. The factors Bewley & Creighton associated with increased risk are identified in Table 17.2. It is worrying that nearly half these "near misses" were related to haemorrhage.
Table 17.2 - High risk features of patients transferred to the Intensive Care Unit; from Bewley (1997). Denominator = 6039 maternities. Note; this Table is not drawn from data in the main Report. Availability of facilities
Cordingley & Rubin carried out a postal survey of all UK obstetric units concerning provision of recovery facilities, HDUs and ICUs in consultant obstetric units7. There was an 89% response rate. Only 62% had a designated staffed recovery unit, 41% had specific HDU beds and there were a number of units without consultant anaesthetic sessions or trained anaesthetic assistants around the clock.
Outcome
The outcome for obstetric patients requiring intensive care has been studied by Lewinsohn et al.8 They found a low standardised mortality ratio (SMR) of 0.416, which is significantly better than that expected. There are various explanations, age alone having been excluded in the analysis - first the subgroup itself may be uniquely different or there may be better care for this subgroup and therefore a better outcome. One explanation offered by Scarpinato & Gerber9 is that the physiological range of variables considered by APACHE II and the weighting for deviation from normal is for a normal, not a pregnant, population and there are changes in physiological variables for the pregnant state (e.g. tachycardia, pH changes, PCV). Another factor may be that APACHE II does not take into account the more serious changes of disease in pregnancy - LFTs, uric acid and platelets. The high emergency caesarean section rate with its significant weighting in the risk equation may have contributed to the findings. Other workers, however, have found that APACHE II, SAPS II & MPMII assess the ICU outcome of critically ill obstetric patients as accurately as non-obstetric patients. References
1. Lapinsky, S.E., Kruczynski, K. & Slutsky, AS. Critical care in the pregnant patient. American Journal Respiratory and Critical Care Medicine 1995; 152: 427-55.
2. Umo-Etuk, J., Lumley, J. & Holdcroft, A. Critically ill parturient women and admission to intensive care: a 5 year review. International Journal of Obstetric Anaesthesia 1996; 5: 79-84.
3. Bouvier-Colle, M.H., Salanave, B., Ancel, P.Y., Varnoux, N., Fernandez, H., Papiernik, E. & Breart, G. Obstetric patients treated in intensive care units and mortality. European Journal of Obstetrics, Gynecology and Reproductive Biology 1996; 65: 121-5.
4. Wheatley, E., Farkas, A. & Watson, D. Obstetric admissions to an intensive therapy unit. International Journal of Obstetric Anaesthesia 1995; 5:221-4.
5. Stephens, I.D. ICU admissions from an obstetric hospital. Canadian Journal Anaesthesia 1991; 38: 677-81.
6. Bewley, S. & Creighton, S. 'Near-miss' obstetric enquiry. Journal of Obstetrics and Gynaecology 1997; 17: 26-9.
7. Cordingley, J.J. & Rubin, A.P. A survey of facilities for high risk women in consultant obstetric units. International Journal of Obstetric Anaesthesia 1997; 6: 56-160.
8. Lewinsohn, G., Herman, A., Leonov, Y. & Klinowski, E. Critically ill obstetrical patients: outcome and predictability. Critical Care Medicine 1994; 22: 1412-14.
9. Scarpinato, L. & Gerber, D. Critically ill obstetrical patients: Outcome and predictability. Critical Care Medicine 1995: 23: 1449-50.
El-Solh, A.A. & Grant B.J.B. A comparison of severity of illness scoring systems for critically ill obstetric patients. Chest 1996; 110: 1299-304.
Lapinsky SE. Critical care management of the obstetric patient. Canadian Journal of Anaesthesia 1997; 44: 325-9.
Lapinsky, S.E. Respiratory care of the critically ill pregnant patient. Current Opinion in Critical Care 1996; 3: 1-6.
Platteau, P., Engelhardt, T., Moodley, J. & Muckart, D.J.J. Obstetric and gynaecological patients in an intensive care unit: A 1 year review. Tropical Doctor 1997; 27: 202-6.
Royal College of Anaesthetists. Guidelines for Purchasers of Obstetric Anaesthetic and Intensive Care Facilities. Revised 1998. London: RCA; 1998.
| ||||
