- Email: [email protected]
Should European perinatal indicators be revisited?
European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
Contents lists available at SciVerse ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Should European perinatal indicators be revisited? Joa˜o V. Santos a,b,*, Cla´udia Correia a, Filipe Cabral a, Joa˜o Bernardes c,d, Altamiro Costa-Pereira a,b, Alberto Freitas a,b a
Department of Health Information and Decision Sciences, Faculty of Medicine, University of Porto, Portugal CINTESIS – Center for Research in Health Technologies and Information Systems, Portugal c Department of Obstetrics and Gynaecology, Faculty of Medicine, University of Porto, Portugal d INEB – Institute of Biomedical Engineering, University of Porto, Portugal b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 December 2012 Received in revised form 11 April 2013 Accepted 27 May 2013
Objectives: Our study presents the results of a survey of physicians and/or researchers working in 21 European countries, on their opinion about the relevance of perinatal indicators, in order to compare it with the EURO-PERISTAT recommendations. Study design: In this cross-sectional study, we selected 21 out of the initial set of 34 indicators of the national data supply on the European Perinatal Health Report, and added four other indicators based on expert opinion. The relative relevance of these 25 perinatal indicators was then rated by 134 respondents – expert physicians and/or researchers who have published in perinatal medicine – through a web-based survey. We summarized our data using descriptive statistics. Results: The top five perinatal indicators, according to the respondents’ rating were: neonatal mortality rate by gestational age, birth weight and plurality; percentage of highly preterm babies delivered in units without a NICU; prevalence of severe maternal morbidity; severe neonatal morbidity among babies at high risk and prevalence of hypoxic-ischemic encephalopathy. Of these top five indicators, however, only neonatal mortality rate by gestational age, birth weight and plurality was considered a core indicator, in 2003. Moreover, severe neonatal morbidity among babies at high risk and prevalence of hypoxicischemic encephalopathy, that were considered in 2003 as requiring further development, were now considered by the respondents as highly relevant. Conclusions: Current views of European physicians and/or researchers working in the perinatal field may not be in agreement with the EURO-PERISTAT recommendations. A revision of the set of perinatal indicators is, therefore, mandatory if a more comprehensive view of health care systems performance across Europe is to be achieved. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Perinatology Quality indicators Health care Europe Health information systems
1. Introduction According to the European Perinatal Health Report [1], despite significant improvements in recent decades, mothers and babies are still at risk during the perinatal period [2]. This period has been defined from the 24th week of gestation until the end of the first week of life, according to the World Health Organization [2], although this is not consensual with the U.S. National Library of Medicine, which defines it from the 28th week of gestation until the 7th day after birth. To improve health care conditions it is important to establish objective quality indicators, in order to analyze the performance of
* Corresponding author at: Department of Health Information and Decision Sciences, Faculty of Medicine, University of Porto, Portugal. Tel.: +351 225513622; fax: +351 225513623. E-mail address: [email protected] (J.V. Santos). 0301-2115/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejogrb.2013.05.017
health systems. Accordingly, we need indicators aiming to assess not only clinical quality and patient care delivery [3] but also social, economic, ethical and physical aspects [4]. As a decision tool, they can therefore be strictly clinical (internal quality) or performance-based (external evaluation) [3]. Clinical indicators need to be measured with little intra- and inter-observer variability, in order to allow generalizations amongst professionals and institutions. An indicator should be, as the event it measures, the most accurate possible, and ensure validity, reliability, measurability and improvability. The clearer and more complete the source material is, more reliability it will achieve [3]. An indicator is useful if it is scientifically supported and important for patient outcome. Furthermore, clinical indicators may allow judgments and setting of priorities, support quality improvement and assist patient choice of providers [5]. The development and measurement of indicators, however, are time consuming [3]. EURO-PERISTAT has a standardized list of perinatal indicators, developed by an expert panel of health professionals with the
86
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
prospect of monitoring and describing perinatal care [6]. EUROPERISTAT’s statistics, however, have not always been based on national common desirable data, but on other possible ones: civil registration, population-based data, hospital-based systems or national surveys, as most countries do not have a mandatory national system to track the quality of delivered care [7]. Moreover, collection of perinatal indicators in each European country depends on its local definition, leading to possible bias when comparing them across Europe [8]. It is known that political interests may be behind that fact [9]. The breadth of information included in the report shows that routine reporting on a wide range of perinatal health indicators is possible in Europe, but it has become clear that practices around Europe differ greatly. Thus a significant effort is still necessary before all European countries are able to contribute to the full set of EURO-PERISTAT indicators [10]. In this setting, our aims were (1) to investigate which perinatal indicators are more relevant, according to experts who have scientific publications related to perinatology; (2) to compare a ranking list, based on respondents’ opinion, with the one available in the European Perinatal Health Report as it dates back to 2003; and (3) to achieve a better and more reliable perinatal collection of data, by the alertness of European countries. 2. Materials and methods Our cross-sectional study was designed to assess the opinion of expert physicians and/or researchers regarding perinatal indicators, by means of a web-based survey. To select them, we first selected journals (Fig. 1) listed in the Journal Citation Report Science Edition 2010 of ISI Web of Knowledge. As inclusion criteria, we used journals belonging to the Obstetrics and Gynecology category, with an ISI impact factor higher than 2.5, or belonging to the Pediatrics category, with an ISI impact factor higher than 1.5. We excluded journals from the Pediatrics category, without an explicit reference in the title to ‘‘Perinatal’’ or ‘‘Neonatal’’. We ended up with 19 journals. From their articles, between January 2009 and August 2011, we selected authors with European
affiliation (using MedLine via PubMed and ISI Web of Knowledge). The absence of available full text or e-mail contact was an exclusion factor. We collected 2059 e-mail contacts, to which we sent a survey using the on-line web-based survey tool MedQuest. All responses were anonymous. The survey included 21 out of an initial set of 34 indicators [7], and excluded the indicators with poor or absent information, supplied to EURO-PERISTAT by national organizations [1]: Prevalence of fecal incontinence, Postpartum depression, Prevalence of late induced abortions, Positive outcomes of pregnancy (births without medical intervention), Neonatal screening policies, Distribution of mother’s country of origin, Maternal mortality ratio by age and mode of delivery, Prevalence of trauma to the perineum, Percentage of all pregnancies following fertility treatment, Prevalence of cerebral palsy, Prevalence of hypoxic-ischemic encephalopathy and Severe neonatal morbidity among babies at high risk (the first six due to the total absence of data). From these, the last four were re-included after an obstetricians’ opinion, who considered them interesting indicators on clinical practice, despite the absence of data available in the report [1]. In this way, we sent a survey with 25 indicators, without any information of its pre-established categories in the European Perinatal Health Report [1]. This could diminish information bias, though each respondent could have that information or could have searched the report. The respondents were asked to evaluate the perinatal indicators as having ‘‘no relevance’’, ‘‘low relevance’’, ‘‘medium relevance’’ or ‘‘high relevance’’, or to select ‘‘do not know’’ if they did not have enough information on the subject. Furthermore, experts were requested to select which indicator was the most relevant in each of the four groups (Group 1: Population characteristics/risk factors; Group 2: Health care services; Group 3: Maternal health; and Group 4: Fetal, neonatal and child health) and to suggest other indicators besides those included in the survey. The respondents were also asked to answer a set of personal questions. Data were extracted from MedQuest and descriptive statistical analysis was performed, summarizing absolute and relative
Fig. 1. Flowchart: description of the process of journal selection.
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89 Table 1 Characteristics of the participant experts (n = 134). Participants characteristics
%
Gender (n = 132) Female Male
47.0 53.0
Head of department (n = 131)
26.0
Occupation (n = 132) Physician Researchers Both Other
57.6 14.4 26.5 1.5
Specialists (n = 108) Maternal care Child care Both Other
45.4 44.4 3.7 6.5
frequencies for each variable. Each indicator relevance class was scored as: ‘‘high relevance’’ – 3 points; ‘‘medium relevance’’ – 2 points; ‘‘low relevance’’ and ‘‘no relevance’’ – 1 point; ‘‘do not know’’ – 0 points. ‘‘Low relevance’’ and ‘‘no relevance’’ were merged owing to their similar importance to derive essential indicators. The criterion to order the indicators by their relevance was the mean score attributed to each one (Table 2). The total sum of the score was also considered, in the event of equal means between two or more indicators. Then, knowing each indicator’s original category – core (C), recommended (R) or further development (FD) – we calculated the average of the mean score of each category. Data were processed using IBM SPSS Statistics 20TM (Statistical Package for Social Sciences). 3. Results We obtained 134 replies out of 2059 inquiries (answer reply rate of 7%). Table 1 summarizes the participants’ characteristics, with a mean age of 47.1 years (SD: 10.2) and an average of 17.0 years of experience (SD: 10.7), according to their specialty: maternal care, if they selected obstetrics or gynecology (49), child care, if they selected pediatrics, neonatology or perinatology (48), or both (4). Fig. 2 displays the 21 European countries from which surveys were returned (data not available for 5 respondents). In Table 2, we present the list of indicators ordered by its mean score. This table shows that not all indicators had the same degree of importance according to specialists’ opinion. Neonatal mortality rate by gestational age, birth weight and plurality; Prevalence of severe maternal morbidity; Percentage of highly preterm babies delivered in units without a NICU; Severe neonatal morbidity among babies at high risk; and Prevalence of hypoxic-ischemic encephalopathy were considered to be the ‘‘top 5’’ indicators. After calculation of the mean score of each category, we obtained the following values: FD – 2.8; C – 2.62; R – 2.42. We also compared all the indicators according to respondents’ age (over 47 vs equal to or under 47, as 47.1 was the mean age of the participants) (Table 1). Interestingly, we found statistically significant differences (p < 0.05) in four indicators: Percentage of infants breast fed at birth; Birth weight distribution by vital status, gestational age and plurality and Infant mortality rate by gestational age, birth weight and plurality were considered more relevant by the older participants, and Prevalence of hypoxic-ischemic encephalopathy was considered more relevant by the younger participants. Moreover, when comparing the geographic origin of the respondents, Neonatal mortality rate by gestational age, birth weight and
87
Table 2 Sum and mean score (ordered by this measure) calculated through class relevance of each indicator: ‘‘high relevance’’ (3 points), ‘‘medium relevance’’ (2 points), ‘‘low or no relevance’’ (1 point), ‘‘do not know’’ (0 points) given by each participant. C: core; R: recommended; FD: Further development (categorized by EUROPERISTAT). Indicator
Category
Mean
Sum
Neonatal mortality rate by gestational age, birth weight and plurality Percentage of highly preterm babies delivered in units without a NICU
C
2.89
379
R
2.84
366
R FD
2.82 2.81
364 357
FD
2.79
351
C
2.77
357
C
2.71
341
C R R R
2.70 2.70 2.69 2.67
356 327 323 352
C C
2.64 2.62
349 341
R C
2.57 2.51
327 321
C
2.48
308
R R R C R
2.36 2.31 2.26 2.23 2.20
312 293 287 292 270
R
2.18
275
R R R
2.09 2.09 2.03
265 259 244
Prevalence of severe maternal morbidity Severe neonatal morbidity among babies at high risk Prevalence of hypoxic-ischemic encephalopathy Fetal mortality rate by gestational age, birth weight and plurality Infant mortality rate by gestational age, birth weight and plurality Multiple birth rate by number of fetuses Prevalence of cerebral palsy Maternal mortality ratio by cause of death Percentage of women who smoke during pregnancy Distribution of maternal age Mode of delivery by parity, plurality, presentation and previous cesarean section Prevalence of selected congenital anomalies Birth weight distribution by vital status, gestational age and plurality Gestational age distribution by vital status and plurality Distribution of mother’s education Perinatal deaths due to congenital anomalies Distribution of Apgar score at 5 min Distribution of parity Distribution of place of birth (according to number of annual deliveries in the maternity unit) Percentage of all pregnancies following fertility treatment Percentage of infants breast fed at birth Distribution of timing of first antenatal visit Distribution of births by mode of onset of labor
Score
plurality and Mode of delivery by parity, plurality, presentation and previous cesarean section were considered more important by Northern Europe respondents. Also, Gestational age distribution by vital status and plurality was considered more relevant by Northern and Western Europe respondents. On the other hand, nonphysician respondents valued Birth weight distribution by vital status, gestational age and plurality; Percentage of infants breast fed at birth; Distribution of births by mode of onset of labor; Prevalence of selected congenital anomalies and Gestational age distribution by vital status and plurality more than physicians. Furthermore, researchers valued Birth weight distribution by vital status, gestational age and plurality; Distribution of births by mode of onset of labor and Distribution of timing of first antenatal visit more than nonresearchers. Finally, female respondents considered Percentage of highly preterm babies delivered in units without a NICU and Prevalence of selected congenital anomalies more relevant than their male counterparts. When the participants were asked to select which indicator was the most relevant in each of the four groups, Distribution of maternal age (31%) was the most chosen in the Population Characteristics/Risk Factors group. In the Health Care Services group, two of the indicators stood out from the others: Percentage of highly preterm babies delivered in units without a NICU (40%) and Mode of delivery by parity, plurality, presentation and previous cesarean section (36%). In the Maternal Health group, with only two options
88
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
Fig. 2. Europe map: depiction of the distribution of the respondents’ nationality.
available, 73% of the participants considered Prevalence of severe maternal morbidity as the most relevant. Lastly, in the Fetal, Neonatal and Child Health group, the most selected indicator was Neonatal mortality rate by gestational age, birth weight and plurality (39%). Some respondents suggested other indicators besides those included in the survey, that can be further developed: Perinatal mortality rate (4 respondents); Long term outcome of the high risk infants (3), Rate of medically assisted deliveries (3), Distribution of maternal social-cultural background (3); Maternal use/abuse of drugs or alcohol (2); Rate of preeclampsia (1); Fetal heart rate distribution (1); Umbilical artery pulsatile index (1); Distribution of mothers by body mass index (1); Rate of asphyxia (1); Number of caregivers involved in prenatal and natal care (number of midwives, residents, GP’s and obstetricians) (1). 4. Comments An indicator of care should represent a measure of quality, for which there is significant evidence, either from scientific research or from a consensus amongst specialists in a particular field or from surveys conducted with patients and/or caregivers [3]. Thus, we developed a survey where physicians and/or researchers were asked to rank perinatal indicators according to their relevance. Efforts were made to improve the predictably low reply rate common in this type of survey, such as resending it to the potential
respondents. Interestingly, we obtained 134 valid answers, a number of participants superior to what is commonly observed in studies with the same methodology. Moreover, as we can see in Table 1 and Fig. 2, the participants represent a large diversity of backgrounds, such as different geographic origins or experience in different fields of expertise. Epidemiologists and other health care scholars may be underrepresented, compared with physicians, but our methods were constructed to extract conclusions from physicians’ professional experience, as this is closer to the handon perspective we intended. Interestingly, comparing the ‘‘top 5’’ perinatal indicators obtained in our study with the EURO-PERISTAT list, only Neonatal mortality rate by gestational age, birth weight and plurality was considered a core indicator, in 2003. Furthermore, indicators such as Prevalence of severe maternal morbidity and Percentage of highly preterm babies delivered in units without a NICU, cited as ‘‘recommended’’ by the report, were considered as highly relevant, as well as Severe neonatal morbidity among babies at high risk and Prevalence of hypoxic-ischemic encephalopathy, previously considered as ‘‘requiring further development’’. These results are in agreement with the ones obtained when the physicians/researchers were asked to select which indicator was the most relevant in each of the four groups. Thus, indicators previously referred assumed a great relevance in their respective groups. After analyzing the disposition of indicators in the ranking, two conclusions can be drawn. Firstly, original ‘‘Further development’’
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
indicators have clearly gained importance as we can also conclude from the average of mean score of 2.80, despite the fact the calculation has only taken into account two indicators. Secondly, ‘‘core’’ indicators have remained highly relevant, with an average of 2.62, compared with ‘‘Recommended’’ indicators, with an average of 2.42. Moreover, some indicators, such as Maternal mortality rate, might be losing relevance in developed countries, as the prevalence of the event decreased to very low values (in this case, 0.1 for every 1000 births) [1,11]. In that way, maternal mortality rates might not reflect the quality of maternal health care – for every maternal death; there are several serious lifethreatening episodes of pregnancy complications, called nearmisses [11–16]. Recent studies [11,14,16] identified maternal near-miss analysis as a valuable contribution to take measures to decrease the likelihood of an avoidable severe complication of pregnancy or childbirth and improve the outcome for women once complications do arise [14]. Also, the combination of severe maternal morbidity and maternal mortality can be a better tool to assess success or failure in perinatal care [11,16]. For some of the new indicators suggested by the respondents, the available literature is scarce. Nonetheless, the suggestions reflect the increasing rates of high-risk pregnancies [17] (as the age of European mothers also increases) and the growing interest in the effect of maternal characteristics (as socio-cultural background influences maternal behaviors) in the pregnancy outcome. Preterm birth is also a major cause of high-risk babies and is also increasing in number [18,19]. The most proposed indicator from respondents was Perinatal mortality rate, suggesting that this indicator might be helpful on mortality rates. Moreover, there may be a need to further investigate if the rising prevalence of caesarian sections [20,21] and other perinatal interventions can significantly interfere, not only with maternal recovery, but also with child development [22]. The main goal of this paper was to acknowledge the perinatal indicators that could better express perinatal healthcare in Europe, according to a specialists’ perspective, in order to standardize perinatal care. We have found a discrepancy between the current views of European experts working in the perinatal field and the recommendations of EURO-PERISTAT. The definition of the core set of indicators may need to be revised or a greater effort is needed in the data collection of perinatal indicators among health institutions and professionals. This is mandatory if a more comprehensive view of European health care systems performance across Europe is to be achieved. Acknowledgments We acknowledge Eng. Jorge Gomes, from CIDES at the Porto Faculty of Medicine, for cooperation on the usage of MedQuest. We
89
also thank Ricardo Santos, MD, for his collaboration in the selection of the perinatal indicators. References [1] Zeitlin J, Mohangoo A. European perinatal health report. Paris: EURO-PERISTAT; 2008, http://www.europeristat.com/images/doc/EPHR/european-perinatal-health-report.pdf (accessed 12.10.12). [2] World Health Organization. Definitions and indicators in family planning – Maternal & child health and reproductive health. World Health Organization. Regional Office for Europe; 2011. [3] Wollersheim H, Hermens R, Hulscher M, et al. Clinical indicators: development and applications. Neth J Med 2007;65:15–22. [4] Reinhardt UE. Quality in consumer-driven health systems. Int J Qual Health Care 1998;10:385–94. [5] Joint Commission on Accreditation of Healthcare Organizations. A guide to performance measurement for hospitals. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2000. [6] Zeitlin J, Wildman K, Bre´art G, et al. Selecting an indicator set for monitoring and evaluating perinatal health in Europe: criteria, methods and results from the PERISTAT project. Eur J Obstet Gynecol Reprod Biol 2003;11:S5–14. [7] Mainz J. Defining and classifying clinical indicators for quality improvement. Int J Qual Health Care 2003;15:523–30. [8] Joseph KS, Liu S, Rouleau J, et al. Influence of definition based versus pragmatic birth registration on international comparisons of perinatal and infant mortality: population based retrospective study. BMJ 2012;344:e746. [9] Kirby RS. International comparisons of perinatal indicators. BMJ 2012;344: e477. [10] Wildman K, Blondel B, Nijhuis J, Defoort P, Bakoula C. European indicators of health care during pregnancy, delivery and the postpartum period. Eur J Obstet Gynecol Reprod Biol 2003;111(Suppl. 1):S53–65. [11] Brace V, Penney G, Hall M. Quantifying severe maternal morbidity: a Scottish population study. BJOG 2004;111:481–4. [12] World Health Organization. International statistical classification of diseases and related health problems, tenth revision (ICD-10). 10th ed. Geneva: World Health Organization; 1992. [13] Salanave B, Bouvier-Colle M, Varnoux N, Alexander S, Macfarlane A, Group MOMS. Classification differences and maternal mortality: a European study. Int J Epidemiol 1999;28:64–9. [14] Lozano R, Wang H, Foreman KJ, et al. Progress towards Millennium Development Goals 4 and 5 on maternal and child mortality: an updated systematic analysis. Lancet 2011;378:1139–65. [15] Wildman K, Bouvier-Colle MH, Group MOMS. Maternal mortality as an indicator of obstetric care in Europe. BJOG 2004;111:164–9. [16] Say L, Souza J, Pattinson R. Maternal near miss – towards a standard tool for monitoring quality of maternal health care. Best Pract Res Clin Obstet Gynaecol 2009;23:287–96. [17] Jolly M, Sebire N, Harris J, Robinson S, Regan L. The risks associated with pregnancy in women aged 35 years or older. Hum Reprod 2000;15: 2433–7. [18] Chang HH, Larson J, Blencowe H, et al. Preventing preterm births: analysis of trends and potential reductions with interventions in 39 countries with high human development index. Lancet 2013;381:223–34. [19] World Health Organization [Internet]. Preterm birth. Fact sheet N8363 [updated 2012 November; cited 2013 March 2]. Available from: http:// www.who.int/mediacentre/factsheets/fs363/en/index.html. [20] Roberts CL, Algert CS, Ford JB, Todd AL, Morris JM. Pathways to a rising caesarean section rate: a population-based cohort study. BMJ Open 2012;2:e001725. [21] Anderson GM. Making sense of rising caesarean section rates. BMJ 2004;329:696–7. [22] Garel M, Lelong N, Marchand A, Kaminski M. Psychosocial consequences of caesarean childbirth: a four-year follow-up study. Early Hum Dev 1990;21:105–14.
Contents lists available at SciVerse ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Should European perinatal indicators be revisited? Joa˜o V. Santos a,b,*, Cla´udia Correia a, Filipe Cabral a, Joa˜o Bernardes c,d, Altamiro Costa-Pereira a,b, Alberto Freitas a,b a
Department of Health Information and Decision Sciences, Faculty of Medicine, University of Porto, Portugal CINTESIS – Center for Research in Health Technologies and Information Systems, Portugal c Department of Obstetrics and Gynaecology, Faculty of Medicine, University of Porto, Portugal d INEB – Institute of Biomedical Engineering, University of Porto, Portugal b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 December 2012 Received in revised form 11 April 2013 Accepted 27 May 2013
Objectives: Our study presents the results of a survey of physicians and/or researchers working in 21 European countries, on their opinion about the relevance of perinatal indicators, in order to compare it with the EURO-PERISTAT recommendations. Study design: In this cross-sectional study, we selected 21 out of the initial set of 34 indicators of the national data supply on the European Perinatal Health Report, and added four other indicators based on expert opinion. The relative relevance of these 25 perinatal indicators was then rated by 134 respondents – expert physicians and/or researchers who have published in perinatal medicine – through a web-based survey. We summarized our data using descriptive statistics. Results: The top five perinatal indicators, according to the respondents’ rating were: neonatal mortality rate by gestational age, birth weight and plurality; percentage of highly preterm babies delivered in units without a NICU; prevalence of severe maternal morbidity; severe neonatal morbidity among babies at high risk and prevalence of hypoxic-ischemic encephalopathy. Of these top five indicators, however, only neonatal mortality rate by gestational age, birth weight and plurality was considered a core indicator, in 2003. Moreover, severe neonatal morbidity among babies at high risk and prevalence of hypoxicischemic encephalopathy, that were considered in 2003 as requiring further development, were now considered by the respondents as highly relevant. Conclusions: Current views of European physicians and/or researchers working in the perinatal field may not be in agreement with the EURO-PERISTAT recommendations. A revision of the set of perinatal indicators is, therefore, mandatory if a more comprehensive view of health care systems performance across Europe is to be achieved. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Perinatology Quality indicators Health care Europe Health information systems
1. Introduction According to the European Perinatal Health Report [1], despite significant improvements in recent decades, mothers and babies are still at risk during the perinatal period [2]. This period has been defined from the 24th week of gestation until the end of the first week of life, according to the World Health Organization [2], although this is not consensual with the U.S. National Library of Medicine, which defines it from the 28th week of gestation until the 7th day after birth. To improve health care conditions it is important to establish objective quality indicators, in order to analyze the performance of
* Corresponding author at: Department of Health Information and Decision Sciences, Faculty of Medicine, University of Porto, Portugal. Tel.: +351 225513622; fax: +351 225513623. E-mail address: [email protected] (J.V. Santos). 0301-2115/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejogrb.2013.05.017
health systems. Accordingly, we need indicators aiming to assess not only clinical quality and patient care delivery [3] but also social, economic, ethical and physical aspects [4]. As a decision tool, they can therefore be strictly clinical (internal quality) or performance-based (external evaluation) [3]. Clinical indicators need to be measured with little intra- and inter-observer variability, in order to allow generalizations amongst professionals and institutions. An indicator should be, as the event it measures, the most accurate possible, and ensure validity, reliability, measurability and improvability. The clearer and more complete the source material is, more reliability it will achieve [3]. An indicator is useful if it is scientifically supported and important for patient outcome. Furthermore, clinical indicators may allow judgments and setting of priorities, support quality improvement and assist patient choice of providers [5]. The development and measurement of indicators, however, are time consuming [3]. EURO-PERISTAT has a standardized list of perinatal indicators, developed by an expert panel of health professionals with the
86
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
prospect of monitoring and describing perinatal care [6]. EUROPERISTAT’s statistics, however, have not always been based on national common desirable data, but on other possible ones: civil registration, population-based data, hospital-based systems or national surveys, as most countries do not have a mandatory national system to track the quality of delivered care [7]. Moreover, collection of perinatal indicators in each European country depends on its local definition, leading to possible bias when comparing them across Europe [8]. It is known that political interests may be behind that fact [9]. The breadth of information included in the report shows that routine reporting on a wide range of perinatal health indicators is possible in Europe, but it has become clear that practices around Europe differ greatly. Thus a significant effort is still necessary before all European countries are able to contribute to the full set of EURO-PERISTAT indicators [10]. In this setting, our aims were (1) to investigate which perinatal indicators are more relevant, according to experts who have scientific publications related to perinatology; (2) to compare a ranking list, based on respondents’ opinion, with the one available in the European Perinatal Health Report as it dates back to 2003; and (3) to achieve a better and more reliable perinatal collection of data, by the alertness of European countries. 2. Materials and methods Our cross-sectional study was designed to assess the opinion of expert physicians and/or researchers regarding perinatal indicators, by means of a web-based survey. To select them, we first selected journals (Fig. 1) listed in the Journal Citation Report Science Edition 2010 of ISI Web of Knowledge. As inclusion criteria, we used journals belonging to the Obstetrics and Gynecology category, with an ISI impact factor higher than 2.5, or belonging to the Pediatrics category, with an ISI impact factor higher than 1.5. We excluded journals from the Pediatrics category, without an explicit reference in the title to ‘‘Perinatal’’ or ‘‘Neonatal’’. We ended up with 19 journals. From their articles, between January 2009 and August 2011, we selected authors with European
affiliation (using MedLine via PubMed and ISI Web of Knowledge). The absence of available full text or e-mail contact was an exclusion factor. We collected 2059 e-mail contacts, to which we sent a survey using the on-line web-based survey tool MedQuest. All responses were anonymous. The survey included 21 out of an initial set of 34 indicators [7], and excluded the indicators with poor or absent information, supplied to EURO-PERISTAT by national organizations [1]: Prevalence of fecal incontinence, Postpartum depression, Prevalence of late induced abortions, Positive outcomes of pregnancy (births without medical intervention), Neonatal screening policies, Distribution of mother’s country of origin, Maternal mortality ratio by age and mode of delivery, Prevalence of trauma to the perineum, Percentage of all pregnancies following fertility treatment, Prevalence of cerebral palsy, Prevalence of hypoxic-ischemic encephalopathy and Severe neonatal morbidity among babies at high risk (the first six due to the total absence of data). From these, the last four were re-included after an obstetricians’ opinion, who considered them interesting indicators on clinical practice, despite the absence of data available in the report [1]. In this way, we sent a survey with 25 indicators, without any information of its pre-established categories in the European Perinatal Health Report [1]. This could diminish information bias, though each respondent could have that information or could have searched the report. The respondents were asked to evaluate the perinatal indicators as having ‘‘no relevance’’, ‘‘low relevance’’, ‘‘medium relevance’’ or ‘‘high relevance’’, or to select ‘‘do not know’’ if they did not have enough information on the subject. Furthermore, experts were requested to select which indicator was the most relevant in each of the four groups (Group 1: Population characteristics/risk factors; Group 2: Health care services; Group 3: Maternal health; and Group 4: Fetal, neonatal and child health) and to suggest other indicators besides those included in the survey. The respondents were also asked to answer a set of personal questions. Data were extracted from MedQuest and descriptive statistical analysis was performed, summarizing absolute and relative
Fig. 1. Flowchart: description of the process of journal selection.
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89 Table 1 Characteristics of the participant experts (n = 134). Participants characteristics
%
Gender (n = 132) Female Male
47.0 53.0
Head of department (n = 131)
26.0
Occupation (n = 132) Physician Researchers Both Other
57.6 14.4 26.5 1.5
Specialists (n = 108) Maternal care Child care Both Other
45.4 44.4 3.7 6.5
frequencies for each variable. Each indicator relevance class was scored as: ‘‘high relevance’’ – 3 points; ‘‘medium relevance’’ – 2 points; ‘‘low relevance’’ and ‘‘no relevance’’ – 1 point; ‘‘do not know’’ – 0 points. ‘‘Low relevance’’ and ‘‘no relevance’’ were merged owing to their similar importance to derive essential indicators. The criterion to order the indicators by their relevance was the mean score attributed to each one (Table 2). The total sum of the score was also considered, in the event of equal means between two or more indicators. Then, knowing each indicator’s original category – core (C), recommended (R) or further development (FD) – we calculated the average of the mean score of each category. Data were processed using IBM SPSS Statistics 20TM (Statistical Package for Social Sciences). 3. Results We obtained 134 replies out of 2059 inquiries (answer reply rate of 7%). Table 1 summarizes the participants’ characteristics, with a mean age of 47.1 years (SD: 10.2) and an average of 17.0 years of experience (SD: 10.7), according to their specialty: maternal care, if they selected obstetrics or gynecology (49), child care, if they selected pediatrics, neonatology or perinatology (48), or both (4). Fig. 2 displays the 21 European countries from which surveys were returned (data not available for 5 respondents). In Table 2, we present the list of indicators ordered by its mean score. This table shows that not all indicators had the same degree of importance according to specialists’ opinion. Neonatal mortality rate by gestational age, birth weight and plurality; Prevalence of severe maternal morbidity; Percentage of highly preterm babies delivered in units without a NICU; Severe neonatal morbidity among babies at high risk; and Prevalence of hypoxic-ischemic encephalopathy were considered to be the ‘‘top 5’’ indicators. After calculation of the mean score of each category, we obtained the following values: FD – 2.8; C – 2.62; R – 2.42. We also compared all the indicators according to respondents’ age (over 47 vs equal to or under 47, as 47.1 was the mean age of the participants) (Table 1). Interestingly, we found statistically significant differences (p < 0.05) in four indicators: Percentage of infants breast fed at birth; Birth weight distribution by vital status, gestational age and plurality and Infant mortality rate by gestational age, birth weight and plurality were considered more relevant by the older participants, and Prevalence of hypoxic-ischemic encephalopathy was considered more relevant by the younger participants. Moreover, when comparing the geographic origin of the respondents, Neonatal mortality rate by gestational age, birth weight and
87
Table 2 Sum and mean score (ordered by this measure) calculated through class relevance of each indicator: ‘‘high relevance’’ (3 points), ‘‘medium relevance’’ (2 points), ‘‘low or no relevance’’ (1 point), ‘‘do not know’’ (0 points) given by each participant. C: core; R: recommended; FD: Further development (categorized by EUROPERISTAT). Indicator
Category
Mean
Sum
Neonatal mortality rate by gestational age, birth weight and plurality Percentage of highly preterm babies delivered in units without a NICU
C
2.89
379
R
2.84
366
R FD
2.82 2.81
364 357
FD
2.79
351
C
2.77
357
C
2.71
341
C R R R
2.70 2.70 2.69 2.67
356 327 323 352
C C
2.64 2.62
349 341
R C
2.57 2.51
327 321
C
2.48
308
R R R C R
2.36 2.31 2.26 2.23 2.20
312 293 287 292 270
R
2.18
275
R R R
2.09 2.09 2.03
265 259 244
Prevalence of severe maternal morbidity Severe neonatal morbidity among babies at high risk Prevalence of hypoxic-ischemic encephalopathy Fetal mortality rate by gestational age, birth weight and plurality Infant mortality rate by gestational age, birth weight and plurality Multiple birth rate by number of fetuses Prevalence of cerebral palsy Maternal mortality ratio by cause of death Percentage of women who smoke during pregnancy Distribution of maternal age Mode of delivery by parity, plurality, presentation and previous cesarean section Prevalence of selected congenital anomalies Birth weight distribution by vital status, gestational age and plurality Gestational age distribution by vital status and plurality Distribution of mother’s education Perinatal deaths due to congenital anomalies Distribution of Apgar score at 5 min Distribution of parity Distribution of place of birth (according to number of annual deliveries in the maternity unit) Percentage of all pregnancies following fertility treatment Percentage of infants breast fed at birth Distribution of timing of first antenatal visit Distribution of births by mode of onset of labor
Score
plurality and Mode of delivery by parity, plurality, presentation and previous cesarean section were considered more important by Northern Europe respondents. Also, Gestational age distribution by vital status and plurality was considered more relevant by Northern and Western Europe respondents. On the other hand, nonphysician respondents valued Birth weight distribution by vital status, gestational age and plurality; Percentage of infants breast fed at birth; Distribution of births by mode of onset of labor; Prevalence of selected congenital anomalies and Gestational age distribution by vital status and plurality more than physicians. Furthermore, researchers valued Birth weight distribution by vital status, gestational age and plurality; Distribution of births by mode of onset of labor and Distribution of timing of first antenatal visit more than nonresearchers. Finally, female respondents considered Percentage of highly preterm babies delivered in units without a NICU and Prevalence of selected congenital anomalies more relevant than their male counterparts. When the participants were asked to select which indicator was the most relevant in each of the four groups, Distribution of maternal age (31%) was the most chosen in the Population Characteristics/Risk Factors group. In the Health Care Services group, two of the indicators stood out from the others: Percentage of highly preterm babies delivered in units without a NICU (40%) and Mode of delivery by parity, plurality, presentation and previous cesarean section (36%). In the Maternal Health group, with only two options
88
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
Fig. 2. Europe map: depiction of the distribution of the respondents’ nationality.
available, 73% of the participants considered Prevalence of severe maternal morbidity as the most relevant. Lastly, in the Fetal, Neonatal and Child Health group, the most selected indicator was Neonatal mortality rate by gestational age, birth weight and plurality (39%). Some respondents suggested other indicators besides those included in the survey, that can be further developed: Perinatal mortality rate (4 respondents); Long term outcome of the high risk infants (3), Rate of medically assisted deliveries (3), Distribution of maternal social-cultural background (3); Maternal use/abuse of drugs or alcohol (2); Rate of preeclampsia (1); Fetal heart rate distribution (1); Umbilical artery pulsatile index (1); Distribution of mothers by body mass index (1); Rate of asphyxia (1); Number of caregivers involved in prenatal and natal care (number of midwives, residents, GP’s and obstetricians) (1). 4. Comments An indicator of care should represent a measure of quality, for which there is significant evidence, either from scientific research or from a consensus amongst specialists in a particular field or from surveys conducted with patients and/or caregivers [3]. Thus, we developed a survey where physicians and/or researchers were asked to rank perinatal indicators according to their relevance. Efforts were made to improve the predictably low reply rate common in this type of survey, such as resending it to the potential
respondents. Interestingly, we obtained 134 valid answers, a number of participants superior to what is commonly observed in studies with the same methodology. Moreover, as we can see in Table 1 and Fig. 2, the participants represent a large diversity of backgrounds, such as different geographic origins or experience in different fields of expertise. Epidemiologists and other health care scholars may be underrepresented, compared with physicians, but our methods were constructed to extract conclusions from physicians’ professional experience, as this is closer to the handon perspective we intended. Interestingly, comparing the ‘‘top 5’’ perinatal indicators obtained in our study with the EURO-PERISTAT list, only Neonatal mortality rate by gestational age, birth weight and plurality was considered a core indicator, in 2003. Furthermore, indicators such as Prevalence of severe maternal morbidity and Percentage of highly preterm babies delivered in units without a NICU, cited as ‘‘recommended’’ by the report, were considered as highly relevant, as well as Severe neonatal morbidity among babies at high risk and Prevalence of hypoxic-ischemic encephalopathy, previously considered as ‘‘requiring further development’’. These results are in agreement with the ones obtained when the physicians/researchers were asked to select which indicator was the most relevant in each of the four groups. Thus, indicators previously referred assumed a great relevance in their respective groups. After analyzing the disposition of indicators in the ranking, two conclusions can be drawn. Firstly, original ‘‘Further development’’
J.V. Santos et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 170 (2013) 85–89
indicators have clearly gained importance as we can also conclude from the average of mean score of 2.80, despite the fact the calculation has only taken into account two indicators. Secondly, ‘‘core’’ indicators have remained highly relevant, with an average of 2.62, compared with ‘‘Recommended’’ indicators, with an average of 2.42. Moreover, some indicators, such as Maternal mortality rate, might be losing relevance in developed countries, as the prevalence of the event decreased to very low values (in this case, 0.1 for every 1000 births) [1,11]. In that way, maternal mortality rates might not reflect the quality of maternal health care – for every maternal death; there are several serious lifethreatening episodes of pregnancy complications, called nearmisses [11–16]. Recent studies [11,14,16] identified maternal near-miss analysis as a valuable contribution to take measures to decrease the likelihood of an avoidable severe complication of pregnancy or childbirth and improve the outcome for women once complications do arise [14]. Also, the combination of severe maternal morbidity and maternal mortality can be a better tool to assess success or failure in perinatal care [11,16]. For some of the new indicators suggested by the respondents, the available literature is scarce. Nonetheless, the suggestions reflect the increasing rates of high-risk pregnancies [17] (as the age of European mothers also increases) and the growing interest in the effect of maternal characteristics (as socio-cultural background influences maternal behaviors) in the pregnancy outcome. Preterm birth is also a major cause of high-risk babies and is also increasing in number [18,19]. The most proposed indicator from respondents was Perinatal mortality rate, suggesting that this indicator might be helpful on mortality rates. Moreover, there may be a need to further investigate if the rising prevalence of caesarian sections [20,21] and other perinatal interventions can significantly interfere, not only with maternal recovery, but also with child development [22]. The main goal of this paper was to acknowledge the perinatal indicators that could better express perinatal healthcare in Europe, according to a specialists’ perspective, in order to standardize perinatal care. We have found a discrepancy between the current views of European experts working in the perinatal field and the recommendations of EURO-PERISTAT. The definition of the core set of indicators may need to be revised or a greater effort is needed in the data collection of perinatal indicators among health institutions and professionals. This is mandatory if a more comprehensive view of European health care systems performance across Europe is to be achieved. Acknowledgments We acknowledge Eng. Jorge Gomes, from CIDES at the Porto Faculty of Medicine, for cooperation on the usage of MedQuest. We
89
also thank Ricardo Santos, MD, for his collaboration in the selection of the perinatal indicators. References [1] Zeitlin J, Mohangoo A. European perinatal health report. Paris: EURO-PERISTAT; 2008, http://www.europeristat.com/images/doc/EPHR/european-perinatal-health-report.pdf (accessed 12.10.12). [2] World Health Organization. Definitions and indicators in family planning – Maternal & child health and reproductive health. World Health Organization. Regional Office for Europe; 2011. [3] Wollersheim H, Hermens R, Hulscher M, et al. Clinical indicators: development and applications. Neth J Med 2007;65:15–22. [4] Reinhardt UE. Quality in consumer-driven health systems. Int J Qual Health Care 1998;10:385–94. [5] Joint Commission on Accreditation of Healthcare Organizations. A guide to performance measurement for hospitals. Oakbrook Terrace, IL: Joint Commission on Accreditation of Healthcare Organizations; 2000. [6] Zeitlin J, Wildman K, Bre´art G, et al. Selecting an indicator set for monitoring and evaluating perinatal health in Europe: criteria, methods and results from the PERISTAT project. Eur J Obstet Gynecol Reprod Biol 2003;11:S5–14. [7] Mainz J. Defining and classifying clinical indicators for quality improvement. Int J Qual Health Care 2003;15:523–30. [8] Joseph KS, Liu S, Rouleau J, et al. Influence of definition based versus pragmatic birth registration on international comparisons of perinatal and infant mortality: population based retrospective study. BMJ 2012;344:e746. [9] Kirby RS. International comparisons of perinatal indicators. BMJ 2012;344: e477. [10] Wildman K, Blondel B, Nijhuis J, Defoort P, Bakoula C. European indicators of health care during pregnancy, delivery and the postpartum period. Eur J Obstet Gynecol Reprod Biol 2003;111(Suppl. 1):S53–65. [11] Brace V, Penney G, Hall M. Quantifying severe maternal morbidity: a Scottish population study. BJOG 2004;111:481–4. [12] World Health Organization. International statistical classification of diseases and related health problems, tenth revision (ICD-10). 10th ed. Geneva: World Health Organization; 1992. [13] Salanave B, Bouvier-Colle M, Varnoux N, Alexander S, Macfarlane A, Group MOMS. Classification differences and maternal mortality: a European study. Int J Epidemiol 1999;28:64–9. [14] Lozano R, Wang H, Foreman KJ, et al. Progress towards Millennium Development Goals 4 and 5 on maternal and child mortality: an updated systematic analysis. Lancet 2011;378:1139–65. [15] Wildman K, Bouvier-Colle MH, Group MOMS. Maternal mortality as an indicator of obstetric care in Europe. BJOG 2004;111:164–9. [16] Say L, Souza J, Pattinson R. Maternal near miss – towards a standard tool for monitoring quality of maternal health care. Best Pract Res Clin Obstet Gynaecol 2009;23:287–96. [17] Jolly M, Sebire N, Harris J, Robinson S, Regan L. The risks associated with pregnancy in women aged 35 years or older. Hum Reprod 2000;15: 2433–7. [18] Chang HH, Larson J, Blencowe H, et al. Preventing preterm births: analysis of trends and potential reductions with interventions in 39 countries with high human development index. Lancet 2013;381:223–34. [19] World Health Organization [Internet]. Preterm birth. Fact sheet N8363 [updated 2012 November; cited 2013 March 2]. Available from: http:// www.who.int/mediacentre/factsheets/fs363/en/index.html. [20] Roberts CL, Algert CS, Ford JB, Todd AL, Morris JM. Pathways to a rising caesarean section rate: a population-based cohort study. BMJ Open 2012;2:e001725. [21] Anderson GM. Making sense of rising caesarean section rates. BMJ 2004;329:696–7. [22] Garel M, Lelong N, Marchand A, Kaminski M. Psychosocial consequences of caesarean childbirth: a four-year follow-up study. Early Hum Dev 1990;21:105–14.