We defined moderate-to-severe perinatal HIE as altered level of consciousness (lethargy, stupor or coma) in the first 6 hours post birth associated with a perinatal history suggestive of a hypoxic-ischaemic insult: pH of cord blood or blood 1 hour post birth of 7 or less, or a base deficit of 16 or greater, and/or a 5-min Apgar score of 5 or less, and/or need of advanced life support.17

We have expressed the incidence rate as the number of live NBs delivered at 35 or more weeks’ gestation with moderate-to-severe HIE relative to the total number of live births reported by the Instituto Nacional de Estadística (National Institute of Statistics). To calculate the incidence relative to the total number of live births at 35 or more weeks’ gestation, we used data provided by the hospitals.

Therapeutic hypothermia: reduction of body temperature to 33°C or 34°C to be maintained for a period of 72h, either: (a) active, attained through manually controlled or servo-controlled devices, or (b) passive, through turning off sources of heat (incubator or radiant mattress) or the use of cold gel pads and/or fans.

We have summarised descriptive data as absolute frequencies and percentages. We have compared data for 2012 and 2013 with the chi square or the Fisher exact tests using the statistical package Epidat 3.1. We defined statistical significance as a p-value of less than 0.05 in any of the tests.

In the 2012–2013 period, 682 NBs delivered at 35 or more weeks’ gestation received a diagnosis of moderate-to-severe HIE (353 in 2012 and 329 in 2013) which, relative to the total 880363 live births in Spain (454648 in 2012 and 425715 in 2013) corresponded to an incidence rate of 0.77/1000 live births (95% CI, 0.72–0.83), with no significant differences between the two years under study: 0.78 in 2012 (95% CI, 0.69–0.86) versus 0.77 in 2013 (95% CI, 0.69–0.86). Fig. 1 and Table 1 show the distribution by AC.

Figure 1.

Incidence of NBs with moderate-to-severe HIE in the 2012–2013 period.

Data expressed as number of moderate-to-severe HIE diagnoses/total live births×1000.

We do not have data for 2012 for Cantabria, so the incidence reported for Cantabria corresponds to 2013.

(0.18MB).

Of the 90 hospitals, 75 reported the number of inborn (delivered at reporting hospital) live NBs of 35 weeks’ or more gestational age, amounting to a total of 372894 live births out of which 406 NBs delivered at 35 or more weeks’ gestational age received a diagnosis of HIE. Based on these data, the incidence of moderate-to-severe HIE was of 1.09 per 1000 live births (95% CI, 0.98–1.20), with no significant difference between the two years under study (1.07 [95% CI, 0.92–1.22] in 2012 versus 1.11 [95% CI, 0.95–1.26] in 2013).

Forty-three percent (290/682) of NBs with moderate-to-severe HIE were born in hospitals currently using TH, with an increase from 36% (126/353) to 50% (164/329) during the period under study (P=.002) (Table 2). All hospitals reported that inter-hospital transport was conducted under passive hypothermia and that in 61% of transfers (55/90), a neonatologist or a team with specialised training was in charge of patient transport (Table 3).

Of the total of 682 NBs with moderate-to-severe HIE, 18% (126/682) died during the neonatal period: 20% (72/353) in 2012 and 16% (54/329) in 2013 (P=.181) (Table 1). These data correspond to a mortality rate due to HIE of 0.14 per 1000 live births (95% CI, 0.12–0.17) in the 2012–2013 period. Table 2 and Fig. 2 show the data broken down by AC, showing the differences between ACs like Catalonia, Andalusia or Aragón, with rates of more than 20%, versus others like Madrid, Castilla-La Mancha, Asturias, Navarre or the Canary Islands, where the proportion of deaths in the population of live NBs with HIE did not exceed 10%.

Figure 2.

Deaths relative to the total number of NBs with moderate-to-sever HIE by AC in the 2012–2013 period.

Data expressed as number of deaths/live NBs with moderate-to-severe HIE

(0.18MB).

The use of TH in the units grew progressively from its introduction in 2008 to its application in 60% (57/90) of tertiary care hospitals by June 2015 (Fig. 3). Ninety-five percent (54/57) of these hospitals used active HT in the form of servo-controlled whole-body cooling. Only one facility had a device for the selective cooling of the head, in addition to the use of whole-body cooling (Table 3). All ACs but 2 had at least one hospital that practiced active TH, and 26% of neonatal units had more than one device to deliver active TH (Table 1) (Fig. 3). The analysis of the availability of active cooling systems by number of live NBs with moderate-to-severe HIE and by number of live births overall showed an uneven distribution, and Navarre and La Rioja were the ACs with the fewest devices per NB with HIE, followed by Murcia and the Balearic Islands (Fig. 4).

Figure 3.

(A) Map showing the 90 hospitals that participated in the survey and their TH capabilities as of June 2015: active TH (solid black circle), passive TH (hollow triangle) or no TH (hollow circle). The autonomous cities of Ceuta and Melilla are not represented because they did not participate in the survey, since none of their hospitals have level III neonatal units and they refer patients requiring this level of care to hospitals in the Iberian peninsula. The progressive increase in the number of hospitals offering TH after its initial introduction in 2 centres in 2008 was of 10 hospitals in 2009, 9 hospitals in 2010, 10 hospitals in 2011, 9 hospitals in 2012, 7 hospitals in 2013, 8 hospitals in 2014 and 2 hospitals in 2015. (B) Distribution of NBs with a moderate-to-severe HIE diagnosis that received TH in the period under study by AC. Dark grey represents children that received active TH, light grey children that received passive TH, and while children that did not receive TH.

(0.42MB).

Figure 4.

Distribution of the number of active cooling systems available as of June 2015 relative to (A) the number of live NBs with moderate-to-severe HIE and (B) the total number of live births; 2012–2013.

In A, data are expressed as number of TH devices/NBs of moderate-severe HIE×1000.

In B, data are expressed as number of TH devices/live births×1000.

(0.28MB).

In the period under study, 86% (588/682) of NBs with moderate-to-severe HIE received TH, a proportion that increased from 84% in 2012 to 88% in 2013 (P=.102), with an increase in the proportion of NBs treated with active cooling from 78% in 2012 (274/353) to 85% in 2013 (279/329) (P=.01) (Table 1). The distribution by AC showed that active cooling was offered the least in the Canary Islands, Castilla-La Mancha and Navarre (Fig. 3). Six percent of the NBs that received TH (35/588) were treated with passive cooling, with the proportion treated with this method decreasing from 8% in 2012 (23/297) to 4% in 2013 (12/291) (P=.089). This decrease corresponded to a reduction by half in the number of NBs treated for no apparent reason with passive cooling as opposed to being transferred to a hospital offering active cooling (Table 1).

Fourteen percent of NBs with HIE (94/682) did not receive TH. If we exclude the 21 NBs that may not have been eligible because they were dying, had pulmonary hypertension and/or a severe clotting disorder, 11% of NBs with HIE (73/682) were not treated with cooling. The most frequent reasons for this were delay in diagnosis or hospital transfer in 31 cases and TH not being prescribed in 20 patients (a number that decreased from 19 in 2012 to 1 in 2013; P=.001) (Table 2).

The incidence rate was consistent with those reported in single-centre studies conducted in Madrid and Barcelona and in a population-based multicentre study conducted in Castilla y León.19–21 The overall incidence of HIE (mild, moderate or severe) in developed countries ranges between 1 and 8 per 1000 live births, and reported rates vary according to the type of study (single-centre, multicentre or population-based) and especially to the operational definition of HIE.22 A review that included 40 studies conducted in 26 European countries and the United States between 1980 and 2013 reported a mean incidence rate of 1.6 per 1000 live NBs delivered at term (range, 0.68–3.75), with the proportion of moderate-to-severe HIE estimated at 60% of the total cases of HIE.1 However, since HIE must be graded in the first hours of life to determine the suitability of TH, we need data on the incidence of HIE at this age. We have not found any other population-based study of moderate-to-severe HIE diagnosed in the first 6 hours of life, information that is essential for the purpose of assessing the health care needs of this population. Our data provide a more accurate picture of the epidemiological situation of HIE, as our study included NBs with HIE delivered between 35 and 37 weeks’ gestation, the population subset currently considered eligible for TH.11,19–21,23

When it came to the distribution by AC, we found considerable homogeneity, the incidence was similar in all of them, although given the nature of the study we cannot be certain that grading was correctly done by all centres. For example, several case have reported that up to 20% of NBs with mild HIE are treated with TH.9,16,23,24 However, it is very likely that the severity criteria applied throughout Spain were very homogeneous, where there are clear national guidelines.17

The use of this therapeutic intervention in Spain has not been introduced following a structured territorial plan, but rather on the initiative of individual neonatologists with the support of the administration of their corresponding hospitals. All participating hospitals had servo-controlled whole-body cooling devices, which have been demonstrated to successfully maintain a stable body temperature without fluctuations.12,25 The proportion of NBs with moderate-to-severe HIE treated with TH (86%) was higher compared to the proportions reported in other countries,10,14 and the use of passive cooling was minimal by the end of the study period (4%). Although good outcomes have been reported with the use of passive TH16 and there are situations where patients are severely unstable and cannot be transported to a hospital equipped to provide active TH, servo-controlled cooling systems offer unquestionable advantages.12,23,25

Another relevant finding in our study was the improvement in the use of TH between 2012 and 2013, which the number of eligible NBs that did not receive this treatment declining from 16% to 12%. This proportion is below those reported in other case series, which can be as high as 35%.10,14,15,26 The reduction in the number of NBs that were not offered this treatment for no apparent reason was particularly significant. Although our study did not allow for the identification of the factors that led to the rapid spread in the use of this treatment, we believe that in addition to the scientific evidence on its efficacy,2,3 the recommendations published by the SENeo17 and the courses or educational programmes specifically focused on the comprehensive management of newborns with HIE20,27 may have contributed to this development.

However, we did not find any improvement in the most frequent reason that HT is not given: a delay in the diagnosis of HIE or in the transfer to a hospital equipped to provide TH.9–11,28 More than half of the NBs with HIE were born in hospitals that did not provide TH, so in order to reduce the number of NBs with moderate-to-severe HIE that do not receive TH in Spain, it is necessary that: (1) hospitals that do not offer TH train their staff to identify NBs with HIE and develop specific protocols for patient stabilisation and safe implementation of passive cooling until the time of transport; transfer in these cases must be considered urgent, but it is possible to delay it a few hours to stabilise the patient, maintaining a stable body temperature of 33–34°C through passive cooling,15,29 and (2) transfer to the referral hospital is carried out under appropriate conditions. Although our study did not analyse the efficacy and safety of patient transport, we found that in a considerable number of instances, transport was carried out by teams without specific training. This underscores the need to have well-trained and specialised teams available in each AC in order to maximise patient safety and maintain body temperature within the therapeutic range, avoiding temperature fluctuations and, above all, overcooling.30,31 An analysis by AC of the travel times and distances from the referring hospital to the receiving hospital may evince the need for portable servo-controlled devices for transport, especially in those ACs where travel times may be long or transfers delayed.

The map of the distribution of hospitals that can provide this treatment in Spain reveals some inequality between ACs and shows those where TH was offered less frequently and should strive to rise to the average standard to improve health care equity. On the other hand, our study also revealed significant differences in the availability of cooling devices. While a large number of devices fitting the number of diagnosed cases of HIE were available in Madrid and Catalonia, there were significant differences between ACs, with some better equipped, like Castilla y León, compared to others like Castilla-La Mancha, Andalusia or Aragón. The differences between ACs in size and in the travel distances to referral hospitals call for alternative solutions, such as collaboration between neighbouring ACs, to develop feasible plans to provide health care to these patients. While another option would be to increase the number of hospitals that offer TH, given the low incidence, complexity of care and need for a multidisciplinary approach in HIE, where TH is but one of the essential elements of care,32 it would be more practical to centralise the management of these infants in hospitals that offer comprehensive multidisciplinary care.7,14,33,34

The mortality due to moderate-to-severe HIE in the 2012–2013 period (18%) was similar to the mortality reported in other case series, which ranges between 15% and 20%,9,10,16,24,26,29 although it was clearly lower than the mortality reported in other studies that estimate it at 27–35%11,19–21 and the 26% mortality reported by the initial clinical trials of TH.2 A recent literature review1 reported an overall neonatal mortality attributable to HIE in countries of similar characteristics to those of Spain of 10% (7.9–11.9%), although the calculation did not include deaths that occurred minutes or hours after birth that are mainly due to perinatal hypoxic-ischaemic episodes. If we were to exclude the group of NBs that die within a few hours of birth from the calculation in our study, the mortality would be 16%, close to the figures reported by Lee et al.1

It is known that most deaths due to HIE take place after end-of-life decisions that include withdrawal of life support.35 Due to the nature of our study, we were unable to determine whether there were differences between hospitals and ACs in this regard or whether the observed differences in mortality were mainly due to differences in HIE severity, but there was considerable variability in mortality between ACs, as some had hospitals reporting mortalities of less than 10% and others hospitals reporting mortalities of more than 20%.

There are some limitations to our study. Although this type of study can provide evidence on actual clinical practice outside the narrow framework of clinical trials,13,23,29,36 its retrospective nature and design (survey) carry a risk of bias, as the collected data may be incomplete or inaccurate. Although there is a possibility that our study underestimated or overestimated the incidence of HIE, we believe this is unlikely, as we obtained responses from every tertiary care centre that may or may not offer TH and that manage NBs with moderate-to-severe HIE, either inborn or transferred from lower-level health care facilities. Furthermore, we carefully reviewed the data submitted by the hospitals that received transferred patients and compared them to the data reported by the referring hospitals in each AC. On the other hand, it is possible that other hospitals have introduced active cooling in the months that elapsed since the survey was completed.

In conclusion, the incidence of moderate-to-severe HIE was similar across ACs. The data of our survey reflect significant progress in the use of TH since its initial introduction in 2008, as nearly 90% of NBs with moderate-to-severe HIE now receive this treatment. Although these findings are encouraging, there are still many areas for improvement, including reducing the unequal distribution of hospitals that can provide TH between ACs, the development of recommendations and specific multicentre programmes for the provision of comprehensive care for HIE involving referring and receiving hospitals and transport teams, and auditing of health care processes and outcomes.