Prospective Risk Factor Monitoring Reduces Intracranial Hemorrhage Rates in Preterm Infants
; ; ; ; ;
Background: Intracranial hemorrhage occurs in 20% to 25% of neonates born before the 30th week of gestation or weighing less than 1500 grams at birth. These hemorrhages carry a risk of long-term neurocognitive damage. Measures for lowering the incidence of intracranial hemorrhage were evaluated.
Methods: A working group at the University of Ulm, Germany, developed a prospective monitoring program for risk factors and a bundle of measures including altered clinical approaches to delivery, initial care of the neonate in the delivery room immediately after birth, and intensive care in the first few days thereafter. Adherence to these measures was checked once per week. The evaluation was performed prospectively for a period of 23 months (August 2010 to July 2012) with a 31-month period of historical controls (January 2008 to July 2010).
Results: In the reference period before the intervention was introduced, 263 neonates weighing less than 1500 grams and with a median (quartile) gestational age at birth of 27.4 (25.4–29.9) weeks were treated. The incidence of intracranial hemorrhage was 22.1%, and that of high-grade hemorrhage was 9.1%. The mortality was 6.1%, and the rate of survival without brain hemorrhage was 74.5%. After the bundle of preventive measures was introduced, 191 neonates weighing less than 1500 grams and with a median (quartile) gestational age at birth of 28.0 (26.0, 30.3) weeks were treated. The incidence of intracranial hemorrhage dropped to 10.5% (odds ratio [OR] 0.43, 95% confidence interval [CI] 0.25–0.73); the incidence of high-grade hemorrhage dropped to 3.7% (OR 0.36; 95% CI 0.14–0.89). The mortality was no different at 6.3%, and 85.3% of the children survived without a hemorrhage (OR 1.95, 95% CI 1.20–3.15). After statistical adjustment for higher gestational age, the OR for intracranial hemorrhage (IVH) was 0.49 (0.28–0.86) and the probability of survival without IVH improved (OR 1.68, 95% CI 1.01–2.81).
Conclusion: The rate of brain hemorrhage in premature neonates can be considerably lowered by prospective monitoring of risk factors.
The survival chances of very premature infants have risen in recent decades (1, 2). This change has been accompanied by an increase in the significance of morbidities in premature infants, including long-term morbidities (3). The risk of brain damage is particularly significant (4, 5). However, the incidence of brain damage has not fallen like that of other morbidities (2). Hemorrhages into the germinal matrix and into the ventricle and hemorrhagic infarctions of the cerebral parenchyma are commonly denoted by the umbrella term “intraventricular hemorrhage” (IVH) (6). Current opinion (6) is that they are caused by reperfusion following ischemia and a range of other factors. They are a specific problem in very premature infants and have significant, long-term neurocognitive consequences (7).
Perinatal units that treat large numbers of cases seem to have a lower IVH rate (8, 9) and mortality rate (10). However, when extremely premature infants are treated, financial constraints obviously also play a role (11). The perinatal unit in Ulm, Germany treats a large number of preterm infants from an area that covers approximately one-fifth of the territory of the region of Baden–Württemberg, as unusually high numbers of pregnant women at risk of preterm delivery are referred there as part of the regional neonatal network (ARGE Ulm) (12). Although comparison of the unit’s neonatal figures with those of Baden–Württemberg as a whole was very favorable, with a mortality rate of 1.44% versus a median standardized mortality rate of 7.15%, the IVH rate was “only” average (13) and was within the range found in the literature (14, 15). Prompted by the experiences of other authors (16, 17), the decision was made to undertake an initiative to reduce the IVH rate at the Ulm perinatal unit.
The goal of this work was to evaluate whether the introduction and prospective monitoring of a bundle of measures that were in line with local procedures and had been developed on the basis of risk factors could reduce the incidence of IVH in preterm infants with a birth weight below 1500 grams (monitoring of quality of process and outcome).
We founded an “IVH Working Group” consisting of the heads of the Departments of Neonatology (Helmut Hummler) and Obstetrics (Frank Reister), senior physicians, interns, and nursing staff. The group met weekly and discussed the pathophysiology of IVH in detail. A delegation of the group visited the hospital that had had the lowest rate of IVH in the federal state for a similar patient and care structure for several years (Heidelberg University Hospital). The hospitals’ treatment standards with regard to individual professional groups were then compared and allocated to one of the following three categories:
- Measures to be adopted directly
- Measures to be examined on the basis of the literature
- Measures not to be adopted
On this basis an individual bundle of measures for IVH prevention that was specific and tailor-made to the Ulm perinatal unit was developed and revised several times. Adherence to the bundle of measures, and thus the quality of the treatment process for all preterm infants with IVH, was discussed at weekly IVH sessions that lasted approximately one hour and brought together various professional groups. By way of “devil’s advocate,” all issues that might have contributed to a hemorrhage were examined. In particular, discussion focused on whether the bundle of measures had been adhered to and whether there were any risk factors.
Ultrasound examination of the head was performed regularly three and seven days after birth, and then at least every four weeks, and the highest grade of hemorrhage according to Papile (18) was recorded. No distinction was made between earlier and later IVH. Mortality was defined as death in hospital.
Statistical evaluation included the data of all infants receiving curative treatment with a birth weight of less than 1500 grams. Data of infants from the cohort born after the bundle of measures was introduced (August 1, 2010 to July 8, 2012) were compared with those of infants born before it had been introduced (January 1, 2008 to July 31, 2010). Infants referred postnatally to Ulm from other pediatric hospitals were excluded, because the approach taken at Ulm could not affect the onset of intracranial hemorrhage in these infants. The study design was that of an interventional cohort study. The dataset before introduction of the bundle of measures was gathered retrospectively; the dataset after its introduction was gathered prospectively.
Statistical evaluation was performed using SigmaPlot 12.0 (Systat Software, San Jose, CA, 2011). Differences in qualitative outcomes were reported using the chi-square test; for quantitative outcomes, the t-test was used to report differences in those with normal distribution, and the Mann–Whitney rank sum test for those with non-parametric distributions. Multivariate logistic regression was performed using SAS 9.2 (SAS Institute, Cary, North Carolina) to investigate differences in the intracranial hemorrhage rate after introduction of the bundle of measures (significance level p = 0.05). The study received a favorable ruling from the Ulm Ethics Committee (no. 213/11).
The IVH prevention bundle included measures concerning delivery, immediate care in the delivery room, and infants’ first few days in the ICU (eBox 1). One of its major component was “minimal handling” (eBox 2), i.e. keeping intervention by physicians and nurses to the absolute minimum. It contained only changes to procedure and cannot be considered in isolation from the pre-existing standard procedure on which it was based (eBox 3). The most important differences it introduced were the following:
- Higher threshold for indicating delivery
- More preference for indicating cesarean section as method of delivery
- Milking the umbilical cord
- An additional single dose of betamethasone
- Timing for the treatment of respiratory distress syndrome and arterial hypotension.
Patient characteristics are summarized in Table 1. Infants included in the cohorts before and after the introduction of the bundle of measures had a median gestational age of 27.4 and 28.0 weeks’ gestation (wks) respectively, and a median birth weight of 910 and 962 grams respectively. The difference in gestational age is statistically significant. There were no significant differences in the following parameters:
- Rate of completed antenatal steroid cycles
- Percentage of infants with dystrophy
- Frequency of bronchopulmonary dysplasia (BPD)
- Discharge home with oxygen
- Necrotizing enterocolitis (NEC)
- Retinopathy of prematurity (ROP).
Despite lower umbilical artery pH values, the cohort before the bundle of measures was introduced had higher Apgar scores. However, subsequently these infants received catecholamines and surfactant more frequently.
Seven infants were born in other maternity units and transferred to Ulm after initial care. Two of these infants suffered intracranial hemorrhages, one before and one after the bundle of measures was introduced. These infants’ data were not evaluated separately, due to the small number of cases.
The incidence of IVH fell from 22.1% to 10.5% (p = 0.002). It decreased in every grade of IVH (Figure), all birth weight categories (Table 2a), and almost all gestational age categories (Table 2b). The fall in IVH incidence was not caused by an increase in mortality: The survival rate remained stable at 94%. The combined target outcome of IVH-free survival was achieved significantly more frequently. Twenty-six infants died, 13 of them less than one week after birth and 10 within one day of birth. Four of the 13 infants had not undergone ultrasound examination of the head, and three were diagnosed with grade III or IV IVH. Of the six remaining infants, one died five days after birth and five died within one day of birth. Five of these deaths occurred before the bundle of measures was introduced, and one afterwards.
The risk profile of the infants changed during the monitored period. The mean gestational age after introduction of the bundle of measures was significantly higher, and the number of infants receiving palliative care in the delivery room after 22 + 0 wks or more rose. Before the bundle of measures was introduced, two infants (0.8%) born after 22 wks received palliative care and died without being admitted to the Division of Neonatology. The corresponding figure after introduction of the bundle of measures was 11 (5.4%) infants, and this difference is significant (+4.7% [95% CI: 1.7 to 7.7%], p = 0.006). If preterm infants with a gestational age of less than 24 wks are excluded, the size of the cohorts is reduced to 237 and 178 infants respectively; the incidence of IVH falls from 18.6% to 10.1% (–8.5% [95% CI: –15.4 to –1.5%], p = 0.024); and the incidence of severe IVH (grade III or IV IVH or death with no ultrasound examination) falls from 7.2% to 3.4% (–3.6% [95% CI: –8.3 to +0.6%], p = 0.145). The combined target outcome of IVH-free survival was achieved in 79.7% of cases before the intervention and 86.5% afterwards (+6.8% [95% CI: –0.6 to +14.1%], p = 0.095).
Multivariate statistical regression was performed to adjust the figures for the higher mean gestational age after the intervention. This reduced the difference in IVH incidence between the two cohorts (Table 3), but the difference in both the IVH rate and the IVH-free survival rate (less than 1500 grams) remained significant.
We conducted an observational study to compare the incidence of IVH in preterm infants with a birth weight below 1500 grams in one cohort established prospectively after the introduction of a bundle of measures to prevent IVH, and one cohort established retrospectively. We observed a 50% drop in the IVH rate in all birth weight and gestational age categories after the bundle of measures was introduced, for all grades of IVH. There were no unclear categorization issues. This success was not achieved at the cost of increased mortality, for example by limiting the care provided to the most seriously ill infants.
Nevertheless, after the bundle of measures was introduced the gestational age of the infants in the study was 0.7 weeks higher, which may have acted as a confounding variable by affecting the incidence of IVH. We therefore performed logistical regression to adjust for the difference in gestational age. As expected, this showed the differences to be smaller (Table 3). It is currently unclear why the mean gestational age was higher after introduction of the bundle of measures. After the bundle was introduced, obstetric care focused more strongly on prolonging pregnancy, whereas previously priority had been given to the prevention of perinatal infections caused by infection of the amnion. However, no increase in perinatal infections was observed. One interpretation of these data is that risk to the fetus and prolongation of pregnancy were weighed against each other more successfully, and pregnant women in whom prolongation of pregnancy would not increase risk were identified.
The reasons for the increase in infants receiving palliative care before 24 wks are unclear. Prenatal counseling of pregnant women at risk of preterm delivery before 24 wks has not changed and complies with Guideline 024–019 of the Association of Scientific Medical Societies in Germany (AWMF, Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften) (19). Parents decide whether treatment should be palliative or curative following non-instructional discussion using facility-specific treatment outcomes. Differences in the aim of treatment for preterm infants born before 24 wks and in the selected reference number may lead to very different results in terms of preterm infants’ survival rate (20). In this cohort, as in most information reported in the literature, the number of live fetuses in utero at 22 wks was not systematically used as the “highest common denominator.” We therefore proceeded as recommended elsewhere (20) and performed a separate evaluation including only preterm infants with a birth weight of less than 1500 grams at 24 + 0 wks or more. This analysis showed lower mortality and IVH incidence than in the total patient population. The effect of the intervention was smaller, but the reduction in the incidence of IVH remained statistically significant.
During the intervention period, the cesarean section rate was significantly higher and surfactant and catecholamine treatment were needed less frequently. There is evidence of an increased risk of hemorrhage for spontaneous delivery (21), particularly in the case of fetal malpresentation (22). Although not all working groups found this relationship (23), we have chosen cesarean section as the method of delivery when the fetus is not in vertex presentation. The decreased need for surfactant was due to higher gestational age rather than to a repeat single dose of betamethasone (24); the route of surfactant administration and indication for ventilation remained unchanged during the period examined. Both lower cardiac output (25) and treatment for arterial hypotension (26, 27) are associated with an increased rate of IVH. We suspect that the reduced need for catecholamines is related to umbilical cord milking (28).
This work does have methodological weaknesses. It included preterm infants who died too early to be able to develop IVH. Therefore, the combined target outcome of IVH-free survival was evaluated. In addition, five of the six deaths without IVH in the first week after birth occurred before the bundle of measures was introduced. If it were assumed that all infants who died would have developed IVH, the incidence of hemorrhage would increase more before the introduction of the bundle of measures than after it, increasing the effect of the intervention. We set the time of initial publication of the bundle of measures as the boundary between the control period and the intervention period even though individual component measures were already known beforehand as a result of the visit to Heidelberg and discussions of the working group. This was also seen in the trend towards decreasing IVH incidence. This means that the decrease in IVH incidence is more likely to have been underestimated than overestimated.
Because of the partially retrospective design of the study, the reason for the drop in IVH incidence cannot be established definitively. Currently ongoing secondary analyses aim to identify the changes that were decisive to success. In other studies, too (16, 17, 29), it proved impossible to identify a single cause for a decrease in IVH incidence. We are inclined to attach more importance to the newly-introduced measures, particularly umbilical cord milking (30) and the additional single dose of betamethasone (31), than to pre-existing treatment aims. However, this is currently speculative and must be tested in detailed analysis.
However, it is also possible that none of the factors mentioned above play a dominant role, but rather that each individual measure has only a small effect and that they only result in success when combined. It is possible, of course, that the activities intended to prevent IVH (study of the literature, visits to another hospital, discussions and introduction of the bundle of measures, weekly sessions) sharpened all team members’ awareness of the pathophysiology of brain damage. IVH is no longer seen as an unavoidable disaster (16) but instead requires critical analysis addressing the question “What should we do differently next time?” It is possible that increased awareness of the problem played the decisive role in reducing the incidence of IVH.
Other quality improvement programs (32), prevention programs (33), and programs for the chronically ill (34) have shown that achieving success and maintaining it are two separate aims. If several measures are combined, they are harder to sustain, particularly if the activities to achieve the effect lapse or competing activities begin (35). Ongoing improvements to the bundle of measures in order to improve quality further are made for this reason alone, but also because of changes in patients’ risk profile.
In order for the success observed here to be transferable, general, transferable elements must be separated from facility-specific elements whose transferability is limited. Facility-specific elements with only limited transferability are the specific points listed in the bundle of measures mentioned here. They are like directions which we have provided for getting from location A (medium IVH incidence) to location B (low IVH incidence). These directions might be completely useless if location A is different. The abstract steps (Key Messages) are transferable.
Brain damage has a substantial impact on the life of affected infants and their families. We believe that this impact justifies major efforts to prevent IVH.
The regional care structure in the context of the regional neonatal network (ARGE Ulm) (12) results in an unusually high number of treated preterm infants with a gestational age of less than 28 wks for the German federal state of Baden–Württemberg. The resulting concentration of specialized expertise, together with interdisciplinary and interprofessional communication with Heidelberg University Gynecology and Pediatric Hospital, may have played a decisive role in this success. In terms of figures, the quality improvement described here prevented 12 intracranial hemorrhages or 5.6 severe intracranial hemorrhages per year when compared to the previous incidence in Ulm Hospital (which was average for the federal region). 9.2 more preterm infants per year survive without intracranial hemorrhage. Success on this scale throughout Germany would save very large numbers of preterm infants and their parents a great deal of suffering. The lifelong costs resulting from mental (36) or physical (37) disability are high and run into the millions; cost-savings would be tremendous.
A prospective intervention in several large perinatal units, for example as part of a cluster randomized trial conducted by a network, would be useful in testing the efficacy of the described measures in a larger population and in reporting the significance of individual factors. At the same time, this might improve the quality of treatment for many preterm infants.
Of the many people who have contributed to this considerable success for the wellbeing of preterm infants, we would like to give particular thanks to Prof. Michael Obladen, who inspired us to do this project; Prof. Johannes Pöschl and the team of the Heidelberg Perinatal Unit for intensive, open discussion; and especially to the nursing staff of Ulm University Hospital, led by Jens Albrecht.
Conflict of interest statement
The authors declare that no conflict of interest exists.
Manuscript received on 2 January 2013, revised version accepted on 22 March 2013.
Translated from the original German by Caroline Devitt, M.A.
Dr. med. Manuel B. Schmid
Sektion Neonatologie und Pädiatrische Intensivmedizin
Universitätsklinik für Kinder- und Jugendmedizin
89075 Ulm, Germany
Department of Obstetrics and Gynecology, Ulm University Medical Center: PD Dr. med. Reister
Institute of Epidemiology, Ulm University: Dr. biol. hum. Mayer
|1.||Horbar JD, Badger GJ, Carpenter JH, F, et al.: Trends in mortality and morbidity for very low birth weight infants, 1991–1999. Pediatrics 2002; 110: 143–51. CrossRef MEDLINE|
|2.||Horbar JD, Carpenter JH, Badger GJ, et al.: Mortality and neonatal morbidity among infants 501 to 1500 grams from 2000 to 2009. Pediatrics 2012; 129: 1019–26. CrossRef MEDLINE|
|3.||Allen MC, Cristofalo EA, Kim C: Outcomes of preterm infants: morbidity replaces mortality. Clin Perinatol 2011; 38: 441–54. CrossRef MEDLINE|
|4.||Sherlock RL, Anderson PJ, Doyle LW, Victorian Infant Collaborative Study Group: Neurodevelopmental sequelae of intraventricular haemorrhage at 8 years of age in a regional cohort of ELBW/very preterm infants. Early Hum Dev 2005; 81: 909–16. CrossRef MEDLINE|
|5.||Volpe JJ: Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol 2009; 8: 110–24. CrossRef MEDLINE|
|6.||Volpe JJ. Intracranial Hemorrhage: Germinal Matrix-Intraventricular Hemorrhage of the Premature Infant. Neurology of the Newborn. 5th ed. Philadelphia, PA: Saunders 2008; 517–88.|
|7.||Mercier CE, Dunn MS, Ferrelli KR, Howard DB, Soll RF, Vermont Oxford Network ELBW Infant Follow-Up Study Group: Neurodevelopmental outcome of extremely low birth weight infants from the Vermont Oxford network: 1998–2003. Neonatology 2010; 97: 329–38. CrossRef MEDLINE PubMed Central|
|8.||Synnes AR, Chien LY, Peliowski A, Baboolal R, Lee SK, Canadian NICU Network: Variations in intraventricular hemorrhage incidence rates among Canadian neonatal intensive care units. J Pediatr 2001; 138: 525–31. CrossRef MEDLINE|
|9.||Synnes AR, Macnab YC, Qiu Z, et al.: Neonatal intensive care unit characteristics affect the incidence of severe intraventricular hemorrhage. Med Care 2006; 44: 754–9. CrossRef MEDLINE|
|10.||Kutschmann M, Bungard S, Kotting J, Trumner A, Fusch C, Veit C: The care of preterm infants with birth weight below 1250 g: risk-adjusted quality benchmarking as part of validating a caseload-based management system. Dtsch Arztebl Int 2012; 109(31–32): 519–26. VOLLTEXT|
|11.||Zimmer K-P: Neonatology departments under economic pressure. Dtsch Arztebl Int 2012; 109(31–32): 517–8. VOLLTEXT|
|12.||Pohlandt F, Artlich A, Freihorst A, et al.: Regionalisation of preterm births in county districts? Yes we can! Z Geburtshilfe Neonatol 2009; 213: 135–7. CrossRef MEDLINE|
|13.||Arbeitskreis der Neonatalerhebungen der Bundesländer: Neonatalerhebung Baden-Württemberg 2009. Zentrum für Qualität und Management im Gesundheitswesen 2010; 1–11.|
|14.||Vogtmann C, Koch R, Gmyrek D, Kaiser A, Friedrich A: Risk-adjusted intraventricular hemorrhage rates in very premature infants: towards quality assurance between neonatal units. Dtsch Arztebl Int 2012; 109 (31–32): 527–33. VOLLTEXT|
|15.||Sheth RD: Trends in incidence and severity of intraventricular hemorrhage. J Child Neurol 1998; 13: 261–4. CrossRef MEDLINE|
|16.||Obladen M, Metze B, Henrich W, Aktas A, Czernik C, Schulz-Baldes A: Interdisciplinary surveillance of intraventricular haemorrhage associated conditions in infants. Acta Paediatr 2008; 97: 731–7. CrossRef MEDLINE|
|17.||Carteaux P, Cohen H, Check J, et al.: Evaluation and development of potentially better practices for the prevention of brain hemorrhage and ischemic brain injury in very low birth weight infants. Pediatrics 2003; 111: e489–96. MEDLINE|
|18.||Papile LA, Burstein J, Burstein R, Koffler H: Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr 1978; 92: 529–34. CrossRef MEDLINE|
|19.||Roll C: Zur Neuauflage der Leitlinie Nr. 024–019 „Frühgeburt an der Grenze der Lebensfähigkeit des Kindes”. Z Geburtshilfe Neonatol 2008; 114–5. CrossRef MEDLINE|
|20.||Smith L, Draper ES, Manktelow BN, Pritchard C, Field DJ: Comparing regional infant death rates: the influence of preterm births. Arch Dis Child Fetal Neonatal Ed 2013; 98: 103–7. CrossRef MEDLINE PubMed Central|
|21.||Leviton A, Fenton T, Kuban KCK, Pagano M: Labor and deliver characteristics and the risk of germinal matrix hemorrhage in low birth weight infants. J Child Neurol 1991; 6: 35–40. CrossRef MEDLINE|
|22.||Shankaran S, Bauer CR, Bain R, Wright LL, Zachary J: Prenatal and perinatal risk and protective factors for neonatal intracranial hemorrhage. National Institute of Child Health and Human Development Neonatal Research Network. Arch Pediatr Adolesc Med 1996; 150: 491–7. CrossRef MEDLINE|
|23.||Malloy MH, Onstad L, Wright E: The effect of cesarean delivery on birth outcome in very low birth weight infants. National Institute of Child Health and Human Development Neonatal Research Network. Obstet Gynecol 1991; 77: 498–503. MEDLINE|
|24.||Peltoniemi OM, Kari MA, Tammela O, et al.: Randomized trial of a single repeat dose of prenatal betamethasone treatment in imminent preterm birth. Pediatrics 2007; 119: 290–8. CrossRef MEDLINE|
|25.||Kluckow M, Evans N: Low superior vena cava flow and intraventricular haemorrhage in preterm infants. Arch Dis Child Fetal Neonatal Ed 2000; 82: 188–94. CrossRef PubMed Central|
|26.||Dempsey EM, Hazzani Al F, Barrington KJ: Permissive hypotension in the extremely low birthweight infant with signs of good perfusion. Arch Dis Child Fetal Neonatal Ed 2009; 94: 241–4. CrossRef MEDLINE|
|27.||Dempsey EM, Barrington KJ: Treating hypotension in the preterm infant: when and with what: a critical and systematic review. J Perinatol 2007; 27: 469–78.CrossRef MEDLINE|
|28.||Hosono S, Mugishima H, Fujita H, et al.: Blood pressure and urine output during the first 120 h of life in infants born at less than 29 weeks' gestation related to umbilical cord milking. Arch Dis Child Fetal Neonatal Ed 2009; 94: 328–31. CrossRef MEDLINE|
|29.||McLendon D, Check J, Carteaux P, et al.: Implementation of potentially better practices for the prevention of brain hemorrhage and ischemic brain injury in very low birth weight infants. Pediatrics 2003; 111: e497–503. MEDLINE|
|30.||Mercer JS, Vohr BR, McGrath MM, Padbury JF, Wallach M, Oh W: Delayed cord clamping in very preterm infants reduces the incidence of intraventricular hemorrhage and late-onset sepsis: a randomized, controlled trial. Pediatrics 2006; 117: 1235–42. CrossRef MEDLINE PubMed Central|
|31.||Vermillion ST, Bland ML, Soper DE. Effectiveness of a rescue dose of antenatal betamethasone after an initial single course. Am J Obstet Gynecol 2001; 185: 1086–9. CrossRef MEDLINE|
|32.||Oretveit J, Staines A: Sustained improvement? Findings from an independent case study of the Jönköping quality program. Quality management in health care 2007; 16: 68–83.|
|33.||Scheirer MA: The life cycle of an innovation: adoption versus discontinuation of the fluoride mouth rinse program in schools. Journal of health and social behavior 1990; 31: 203–15. CrossRef MEDLINE|
|34.||Reinehr T, Widhalm K, L'Allemand D, et al.: Two-year follow-up in 21,784 overweight children and adolescents with lifestyle intervention. Obesity (Silver Spring) 2009; 17: 1196–9.|
|35.||Wiltsey Stirman S, Kimberly J, Cook N, Calloway A, Castro F, Charns M: The sustainability of new programs and innovations: a review of the empirical literature and recommendations for future research. Implementation science: IS BioMed Central Ltd 2012; 7: 17.|
|36.||Centers for Disease Control and Prevention (CDC): Economic costs associated with mental retardation, cerebral palsy, hearing loss, and vision impairment – United States, 2003. MMWR – Morbidity & Mortality Weekly Report 2004; 53: 57–9. MEDLINE|
|37.||Kruse M, Michelsen SI, Flachs EM, Brønnum-Hansen H, Madsen M, Uldall P: Lifetime costs of cerebral palsy. Dev Med Child Neurol 2009; 51: 622–8. CrossRef MEDLINE|