OUP user menu

From knowledge to planning considerations: a matrix to assess health needs for the perinatal network in eastern Paris

Annie Serfaty, Francis Gold, Jean-Louis Benifla, Gérard Breart
DOI: http://dx.doi.org/10.1093/eurpub/ckq101 504-511 First published online: 31 July 2010

Abstract

Background: Regionalization of perinatal care is required throughout networks to reduce perinatal morbidity and mortality and to organize access to health services for high-risk new borns, such as in the eastern Paris in 2007. Our study sought to design a matrix to build a perinatal knowledge base for assessing health needs and facilitating public health planning process for the perinatal network in eastern Paris. Methods: Our matrix listed as its columns ‘perinatal stages’ from conception through the age of 6 years, whereas the rows covered components related to public health planning (i.e. target population, perinatal risk factors and health services). For each situation, the matrix lists require information and potential data sources to measure health status and health services. Results: Our matrix structures the cyclical process for building knowledge for action. The eastern Paris has a population of 670 000. Its nine maternity units are distributed into three levels of care, a level-3 unit opened up in June 2007. A total of 16 400 deliveries occur every year in the nine units; 2500 women living in the eastern area deliver outside the area. Conclusion: Our matrix is useful for building a comprehensive perinatal knowledge base depending upon perinatal stages and health-care dimensions. It should imply stakeholders in collecting, synthesizing and analysing massive amounts of data. It can be adapted to any health network or local health policy.

  • health needs assessment
  • health services research
  • network
  • perinatal care
  • regionalization

Introduction

Since 1970, a series of official plans and government programmes have addressed to improve perinatal health in France.1,2 The 1994–2000 plan3 introduced regionalization of perinatal care as its underlying strategy to reduce perinatal morbidity and mortality and to organize access to health services for high-risk new borns.4 Various evaluations5–12 provide scientific knowledge on the effect of place of delivery on survival of very preterm babies. Official regulations were enacted in 199813 to necessitate that each maternity unit be accredited as a level 1, 2 or 3, depending upon the equipments and level of neonatal services and intensive care it provides.

The high density of maternity units in the Paris area resulted in postponing the implementation of regionalization based on health areas.14,15 A level-3 perinatal centre, a maternity unit with on-site neonatal intensive care unit (NICU), was implemented in June 2007, in the paediatric hospital within the Paris public hospital system (AP-HP). It must provide facilities for very preterm births and high-risk deliveries for the population (670 000 inhabitants) of four arrondissements (11th, 12th, 13th and 20th) in eastern Paris.14

Construction of a health network began with an aim to ensure that regionalization and an appropriate range of perinatal facilities would meet the needs of this population.14 Health needs must be defined in conjunction, on one hand with the key values in the recent plan (2005–07)16—(i) to guarantee treatment that is humane, nearby, safe, and of high quality for mother and newborn; (ii) to improve the medical environment of pregnancy and delivery; (iii) to improve the psychosocial environment of mother and child; and (iv) to modernize practices and training for perinatal professionals; and on the other hand with the objectives of the government plan to reduce perinatal and maternal morbidity and mortality.

The law enacted on 4 March 200217 defined the objectives of health networks: ‘to ensure suitability of health-care service delivery in terms of health education, prevention, diagnosis, and treatment for the population residing in a geographic area; and to improve coordination, continuity, and interdisciplinary health care for the target population’. This population-based approach for reducing morbidity and mortality in both mothers and babies by providing accessible and appropriate services for prevention and care requires the collection and sharing of information among stakeholders. The construction of such a perinatal public health knowledge base to assess public health needs requires dealing with a variety of data sources that are not yet available at the local level and integrated into an ‘interoperable’ information system.18

We thus sought to design a matrix that would enable searching, collecting, managing, synthesizing and analysing a massive amount of data to be included in the collective knowledge base, which is vital for assessing perinatal health needs in eastern Paris. This article presents the concept of this matrix.

Methods

Following the definition of a health network, as ‘a structured and organized collective action’,19 we consider the perinatal health network in eastern Paris to be a part of the collective action of building a public health policy.

We propose to combine a public health policy approach with evidence-based public health (figure 1). Our conceptual framework is based on: (i) a five-step approach to public policy, as outlined by C.O. Jones,20 which includes: programme identification, development, implementation, evaluation and termination; and (ii) an evidence-based public health approach.21 The statement of the issue includes a complete description of the problem, potential solutions, data sources and health-related outcomes.21,22 It represents the initial strategic stage of the health planning process. Here it corresponds to the construction of a perinatal public health knowledge base to assess perinatal health needs for the target population of the network.22–24

Figure 1

A sequential planning cycle for building a Public Health knowledge base for a perinatal network. The framework combines public health policy approach and evidence-based public health, adapted from: a sequential framework for enhancing evidence-based public health21

To construct the perinatal public health knowledge base, our matrix listed as its columns the various ‘perinatal stages’ from conception throughout the age of 6 years, whereas the rows covered the components related to public health planning (i.e. characteristics of the target population; health risk factors to be taken into account to determine appropriate services; existing preventive health services; health-care delivery; professional practices; user demands and/or satisfaction). These various dimensions can then be integrated to search, collect, manage, synthesize and analyse a huge assortment of data within the perinatal domain.

To assess perinatal public health needs for the population of eastern Paris, we compared the information collected with reference data from national, European or international sources to identify discrepancies between what is observed and what should be reached in terms of perinatal health indicators.

Results

Matrix to build a perinatal public health knowledge base

As illustrated in table 1, the goal of our matrix is to structure the cyclical process of constructing a perinatal knowledge base to assess health needs for the perinatal network in eastern Paris. This process is fundamental to ensure that the objectives and interventions are appropriate to the needs of this specific population and to follow health outcomes.

View this table:
Table 1

Matrix to build a public health perinatal knowledge base for the health network in eastern Paris (2008–10): from needs assessment to outcomes evaluation

Perinatal stagesBefore conceptionPregnancyDeliveryBirthPostpartum follow-up of mother and baby
Components for health planningRequired dataPotential data sourcesRequired dataPotential data sourcesRequired dataPotential data sourcesRequired dataPotential data sourcesRequired dataPotential data sources
Target populationGeneral population: structure and trends; childbearing women; births; fertility; infant deaths; women or couples with infertility problemsINSEE: population census; civil registration on birth certificates; civil registration on death certificates; local social agency data sources; Register of assisted reproductive techniques (ART) procedurePregnant women = women who gave birth + women with miscarriage + women with an ectopic pregnancy +  women who have abortion + women with a termination of pregnancy; data on women living in the area and/or women using health services in the areaINSEE: civil registry on birth certificates; annual public and private hospital activity data; hospital discharge data systems; National Institute for Demographic Studies (INED)Women who deliver in the maternity units of the area, and those who live in the area; home deliveryAnnual public and private hospital activity data; hospital discharge data system; National Birth certificateLive births including multiple births, very preterm births, low weight birth, longer term; stillbirths, infant deaths; neonatal deaths; births from mother living in the territory, and those occurred in maternity units in the area; home birthsINSEE: civil registration on birth certificates; civil registration on death certificates; vital registry data; annual public and private hospital activity data; hospital discharge data system; EPIPAGE (INSERM U 953, ex U-149)Mothers and children using facilities of the area and those living in the territory; live births and high-risk babies; congenital anomalies; women and babies with medical or social risk; disabled infantsVital registry data (INSEE); annual public and private hospital activity data; hospital discharge data system; EPIPAGE (INSERM U 953, exU-149); Ministry of Health, Maternal and child protection: children's health certificates (within 8 days, at 9 months, at 24 months)
Risk factors and health determinantsMarital status; educational level of childbearing women; social characteristics (lower social status, population receiving social benefits); mother's age at birth; alcohol consumption, tobacco use, substance abuse; environment quality; pathologies (hypertension, diabetes, obesity, etc); unwanted or hidden pregnanciesINSEE: population census, civil registration on birth certificates, civil registration on death certificates; National Perinatal Survey; newborn health certificate (CS8); INED; French Decennial Health SurveyMother's age at birth; parity; existing disease (hypertension, diabetes, obesity, heart disease, etc); alcohol consumption; maternal smoking; unfavourable socio-economic conditions (median income, % receiving welfare/income assistance, immigrant status); mother's educational levelINSEE: population census, civil registration on birth certificates, civil registration on death certificates; National Perinatal Survey; Newborn health certificate (CS8); hospital discharge data system; social family support dataMaternal age, parity, plurality (singleton, twin or triplet pregnancies); lack of antenatal care; rate of women who had less than seven antenatal visits; first antenatal visit; pregnancy complications; delivery; complications; delivery by caesarean section; very preterm deliveries; stillbirth; place of delivery; maternity unit inappropriate for maternal and/or foetal risks; woman not known at maternity unitCivil registration on birth certificates; civil registration on death certificates; National medical hospital discharge data systems; death certificates (INSERM-CepiDC); National Perinatal Survey; newborn health certificate (CS8)Multiple births; preterm births; low birth weight; congenital anomalies; foetal alcohol syndrome; delivery before 33 weeks and birth weight <1500 g; place of birthAnnual public and private hospital activity data; hospital discharge data systems; Registry of congenital anomalies (Paris); EPIPAGE (INSERM U 953, ex- U 149); newborn health certificate (CS8)Socio-economic status; family housing status; maternal age, parity, plurality (singleton, twin or triplet or higher order pregnancies); mother's mental healthINSEE: population census, social family support data (data from tax authorities INSEE); National Perinatal Survey
Preventive services and actionsSex education at middle/high school interventions; family planning services; intervention provided by non-governmental organizations; number of prevention sites available for the population; cessation smoking programmeDirectory of the Town Hall; local health administration; family planningPublic/private maternity units; health-care professionals (employed or private practice); antenatal care visits; prenatal diagnostic centres; trisomy 21 prenatal detection; early interview for risk screening; access to preparation for delivery/birth; health telephone hot line; support information and counselling on breastfeedingAnnual public and private hospital activity data; local health administrationPublic/private maternity units; supply of neonatology services; transporters/ambulancesAnnual public and private hospital activity data; local health administration; SAMU-SMUR (mobile emergency medical services)Public/private maternity units and neonatology services; breastfeeding support; evoked otoacoustic screeningAnnual public and private hospital activity data; local health administration; SAMU-SMUR (mobile emergency medical services)Maternal and child protection programmes; physician in private practice; non-governmental organizationAnnual public and private hospital activity data; local heath administration
Health-care facilitiesHealth centres; maternal and child protection; private professionals (general practitioners, gynaecologists, etc); ART sitesYellow Pages; private medical practice; observatory; annual health facility activity data; professional recordPublic/private maternity units;health-care professionals (employed, private); antenatal care visits; prenatal centres; laboratories, ultrasonographersAnnual public and private hospital activity dataObstetric units, neonatology units, neonatal intensive care units (number of authorized and actual beds); sites for arterial embolization; access to labile blood products (procedures)Annual public and private hospital activity data; regional hemovigilan-ce committeePaediatrician in maternity units; neonatal units and neonatal intensive care units (number of beds)Annual public and private hospital activity dataHome hospitalization; hospital consultations; private consultations; services delivery for high-risk childrenAnnual public and private hospital activity data; SAMU-SMUR (mobile emergency medical services)
Professional practicesInformation and prevention practicesObservatory of health rightsProportion of physicians who provide appropriate information and prevention advicesObservatory of health rights; assessment of professional practices (National Authority of Health, hospital, clinics)Prenatal transfers; access to appropriate maternity unit; rate of caesarean section; instrumental deliveryPublic and private hospital activity data; hospital discharge data system; mortality and morbidity staff; reviewsNewborns and mothers in the same hospital site; access to appropriate birth site; place of birth for preterm babiesHospital discharge data system; specific recordsClear and fair information provided; adequate information on breastfeeding; referral to the appropriate departmentObservatory of health rights; assessment of professional practices
Stakeholders inputHealth-care professionals (employed or in private practice); those involved in programmes operation; coalition partners; administrators and staff; health usersNetwork registrationHealth-care professionals (employed or in private practice); those involved in programmes operation; coalition partners; administrators and staff; health usersNetwork registrationHealth-care professionals (employed or in private practice); those involved in programmes operation; coalition partners; administrators and staff; health usersNetwork registrationHealth professionals (employed, in private practice); involved in programmes operation; coalition partners; administrators; health usersNetwork registrationHealth-care professionals (employed or in private practice); those involved in programmes operation; coalition partners; administrators and staff; health usersNetwork registration
Potential perinatal outcomesRate of preconception counselling among the childbearing womenProgramme recordsRate of antenatal visit care among low income pregnant womenProgramme records; newborn health certificate (CS8); National Perinatal SurveyRate of high-risk and preterm deliveries in level-3 unit; rate of multiple delivery in level 1, 2 or 3 unit; postpartum haemorrhages; maternal deathsAnnual public and private hospital activity data; hospital discharge data systemHigh-risk and preterm births in level-3 unit, in units without a NICU; multiple delivery in level 1, 2 or 3 units; breastfeeding at birthAnnual public and private hospital activity data; hospital discharge data systemMaternal mortality and morbidity related to childbearing; breastfeeding rateINSEE data

For each situation, i.e. a ‘perinatal stage’ combined to a public health planning component, the matrix lists the required specific information and the potential data source. Thus making it possible to conceptualize the health policy process on the one hand, and to provide information and feedback on public health interventions to sustain stakeholder mobilization, on the other. Our matrix includes three important components: a knowledge base and information gathering on health status; prevention and health-care services; and health outcomes.

Target population for the eastern Paris perinatal network

As the matrix shows, the target population is defined depending upon the perinatal period considered and various data sources required. Demographic data, i.e. general population, age distribution, births and mothers’ ages at birth, can be obtained from the French National Statistical Institute (INSEE),25 and the disadvantaged population receiving social benefits can be identified through local social agency data sources.

For the study of period before conception, the target population comprises the childbearing women (15–49 years of age), who account for 29% of the population (670 000 inhabitants). One of the most important services that the network must furnish is information about infertility treatments, contraceptions, prenatal care, delivery and postnatal care. Information about perinatal health-care facilities, in the area of the network, is essential to guide each woman and couple to a facility appropriate to their level of risk. The distribution of information through a public campaign is the basic method of community education for the target population of the area. In order to estimate the number of leaflets necessary, we need to know the number of women in the designated age categories. Specific barriers must also be taken into account, such as language differences in people born outside of France and lack of knowledge about accessibility of health-care services. In order to reach women from different nationalities and cultures, ∼20% of the leaflets will need to be translated into various languages and be distributed through different women’s organizations. This is an example of service provision based on knowledge.

During pregnancy through delivery and birth, the concerned population is the women who become pregnant. In order to estimate this population, we must count deliveries occurring in the nine maternity units in the health area, the women living in the area who give birth outside the area, and the spontaneous, elective and medically indicated abortions. Different data sources must be analysed: population census and civil registration on birth statistics25 that give information on births among the resident population, statistics from public and private hospital activity data26 and hospital discharge data systems27 that identify deliveries from standardized records.

At the delivery/birth stage, the concerned population is, on the one hand, the women who deliver in the nine maternity facilities in the area, as well as those women who live in the area and give birth elsewhere, and, on the other hand, the newborn receiving care in area maternity units and other children living in the area. Information can be found from public and private hospital activity data26 as well as from the hospital discharge data systems.27 In assessing health-care needs, it is important to combine a population-based approach with a health services utilization approach. In 2008, this area included nine maternity units that accounted 17 112 deliveries and 17 400 births; almost 2500 women living in the eastern area delivered in a maternity unit outside the area.

During the postnatal period and throughout childhood (to the age of 6 years), the target population comprises the mothers and children living in the area and those using services from the area. They very often continue to use obstetrical, paediatric and other medical services participating in the perinatal network.

Maternity units and neonatal services

The distribution of the maternity units in any health area must be presented according to their designated level of perinatal care, defined by the neonatal care facilities it provides.14 Annual statistics from the public and private hospital activity data26 and administrative decision records are the principal data sources to describe specific equipment, personnel and care activities. Since June 2007, nine maternity units are implemented in the eastern area of Paris: five are level 1 (maternity unit without on-site neonatal service), three are level 2 (maternity with on-site neonatal care) and one is level 3 (maternity with an on-site NICU). Also, 52 beds are counted for neonatal medical care, 12 in intermediate NICU and 8 in full NICU.

To assess perinatal health-care equipment needs, we must compare the services available and provided with the perinatal services needed and thus plan an appropriate distribution of each level of care. For example, for the neonatal equipment, we can apply the national criteria28,29 to measure the required number of beds in neonatal medical care (2–3 beds for 1000 births), in intermediate NICU (1–2 beds), and in full NICU (0.5–1.5 beds). According to the 17 400 births occurring on the eastern Paris, the estimation of beds is: from 35 to 52 beds for neonatal medical care, 17–35 for intermediate NICU, and from 9 to 17 for full NICU. Thus, we can measure discrepancy between existing neonatal beds and beds that are required depending upon national criteria in the health area: a lack of intermediate NICU and full NICU beds to meet population needs.

Discussion

The matrix proposed here supports the development of a public health knowledge base to assess health needs for the perinatal network and subsequently plan implementation of interventions in eastern Paris. It is intended to structure the discussions with the various stakeholders about perinatal health knowledge in the eastern area, shared objectives and intended perinatal outcomes. It can be applied to any public health policy at a local or regional level.

One of the principles that guided the conceptualization of this matrix is that identification of population needs makes professional mobilization possible and thus builds bridges between public health knowledge and services.30 We chose to assess public-health needs as part of a policy-making process that is both population-based and health-care facility-based and that combines the mobilization of stakeholders with formal planning, including data collection, the definition of objectives, the provision of services, and interventions and their evaluation.22

This knowledge base requires information from different sectors, i.e. social and demographic data, health risks and determinants, prevention services and health-care facilities, professional practices, consumer satisfaction, and perinatal indicators. It demonstrates the need to formalize a perinatal health information system that links data production with the analysis and synthesis of existing data. This framework thus combines different existing data sets—based on the entire population for births, on the population using prevention and health services, and on other sources, such as national studies of representative samples, such as the French National Perinatal Survey.31,32 Finally, these data are combined with the PERISTAT perinatal indicators,33 specifically required for evaluating perinatal networks in France.34 The European PERISTAT project’s scientific committee selected and defined, through a consensus process, indicators that are considered important for monitoring and evaluating perinatal health in the European Union: 10 core indicators, recommended indicators and those for further development.35 Nevertheless, PERISTAT indicators are not sufficient to assess public health needs and measure health outcomes for a network. Other information must be integrated, including the mobilization of stakeholders and indicators for prevention services and health-care facilities, such as the number of obstetric and/or neonatal beds for the population, whether defined according to the area or according to service utilization.

Most of the information for the eastern Paris area was integrated into national data systems. For example, the number of births according to mother’s residence can be determined from the population census, but not the births at a particular maternity unit. We need to combine demographic data, public and private hospital activity data, and data from hospital discharge system. These sources collect information specific to their own needs, with different units of analysis and different definitions, i.e. population census, hospital activities and health-care reimbursements. For example, some count the number of deliveries and births, whereas others count hospitalizations for delivery. In those cases, newborns are linked with their mothers and not considered independent patients, except when they are admitted to a neonatal care unit. The data sources report different numbers of births, possibly explained by whether the unit recorded is the birth, the woman giving birth or the hospitalization admission. Moreover, depending on the source, the delay between collection, production and diffusion of data ranges from 1 to 4 years. The Ile-de-France Regional Hospital Agency builds a perinatal information system based on the standardized hospital discharge records to support perinatal information networks.27

The underlying questions remain: How can we take into account the magnitude of the existing data? How can we assess the degree of significance and interactions within the information needed to build public health knowledge base? Effectively managing overflow information is the challenge of the new millennium. Much existing information has not yet been analysed and synthesized.36

An important stage of the planning process is the definition of the target population that health needs are to be assessed. As we observed above, the definition of the target population for the network must take into account the perinatal period, as we defined it in our matrix. For the period of pregnancy and delivery, two approaches must be combined: one based on the resident population and one based on health service users. We utilized three data sources: civil registry birth data,25 hospital and clinic discharge records,26 and annual public and private hospital activity data.27 The difference in the number of births to residents (9216) and births in the network facilities (17 400) raises numerous questions: Does the location where the mother works explain this difference? Does the second parent’s work location is situated near the maternity unit? Does the distribution of the maternity units in the adjoining health areas explain it? One perspective is to follow behaviours of the pregnant women in the eastern area, to analyse how attractive a level-3 perinatal unit is to women and their families. We must also determine the extent to which socio-economic characteristics affect selection or accessibility to health-care services.

The fundamental question is how a perinatal network can assess the needs for prevention and health services of an area, and reach the population’s needs within an open geographic system. The population mobility is a factor to be taken into account as part of the human condition.

Although we know that perinatal health services and interventions must be oriented according to medical and psychosocial risks, data on risk factors and determinants is difficult to obtain routinely at the local level. Moreover, defining unmet needs will probably conduct stakeholders to define public-health actions towards prevention of risk factors to reduce inequalities.22,37 We will develop these aspects in a future stage of our research. Health risk data can be documented from the periodic national perinatal surveys conducted in France in 2003, which produce estimates of such indicators as the number of women who did not have their seven recommended prenatal visits and of other risk factors (e.g. low educational level, lower socio-economic factors).31,32,38 The mandatory newborn certificate, to be completed between birth and the eighth day after birth, includes reported health and socio-economic risk factors for babies. The epidemiological data produced by this certificate is under question.39 Nonetheless, there are available indicators of social well-being that reflect multiple social determinants of health. One such example is infant mortality, i.e. deaths in the first year of life.40

Some data on clinic and hospital equipments are available, such as the number of neonatal or obstetric beds. However, more information is required to evaluate other resources and services, such as number of staff professionals, and also quality and security of care. Moreover, it is difficult to define indicators that link theoretical norms to effective services, health security and easy access to optimal care. A study conducted by Gouyon et al. pointed out how services may be mismatched: 59% of children received an unnecessarily high level of services and 21% an insufficient level (i.e. required and did not receive either intermediate or full NICU).41

Place of birth for babies <1500 g birth weight, rate of multiple births and caesarean deliveries are indicators that should be documented each year, for evaluating perinatal regionalization and professional practices. Place of births for preterm births and multiple pregnancy are indicators to assess regionalization in a perinatal network.31,35,42 An increase of caesarean rate suggests excessive medicalization and inappropriate adjustment of health-care delivery to medical and psychosocial risks of mother and child.43 A comprehensive system with key indicators will be needed to assess the services provided by the perinatal health network, to link professional interests and practices to user satisfaction, and to develop an annual report that justifies adequate financial support from the regional health administration.

Our matrix seeks to construct a perinatal public health knowledge base to facilitate the health needs assessment and the definition of public health objectives for a specific area of a health network. It participates to develop the links between health needs assessment and stakeholders input, that are necessary to reach a consensus about public health objectives and reducing inequalities in accessing health services of a network.

The results of our research are presented as a model for creating and structuring bridges from knowledge to action, including political will and stakeholders input. This model must be applied in integrating each element in an information chain, i.e. data collection, data analysis and dissemination of findings towards stakeholders and health users, to create the initial issue of the statement that are required for health network needs assessment, and prevention and health services planning process.

Acknowledgements

This study was made possible through the tutorial and the supervision received from the team of the Research Unit on Perinatal Health and Women’s Health, UMR S 953 (ex-U149), INSERM, Paris.

An oral presentation was made in October 2008, at the 38ème Journées Nationales de la Société française de Médecine Périnatale. Strasbourg (FRANCE): Serfaty A, Gold F, Bénifla JL, Bréart G. Le Réseau Périnatalité de l’Est Parisien. Définition de la population et évaluation des besoins en 2008. The Eastern Paris Perinatal Network (2008–2010) Defining the target population and assessing health needs.

Conflicts of interest: None declared.

Key points

Our matrix seeks to build a perinatal public health knowledge base, model that can be adapted to any health network. This matrix is an important step in an evidence-based planning process for a local policy:

  • In structuring a systemic approach to collect, manage and analyse a massive amount of data to be able to document perinatal indicators, depending upon perinatal stages, for the mother and the new born, professionals concern with practice evaluation, and stakeholder implication;

  • In managing and structuring the planning process of a local health policy;

  • In building bridges between knowledge, planning, implementation, evaluation and stakeholder input.

An implication for public health policy and practice is to develop a model with a systematic approach in integrating the dimension of public health knowledge for assessing health needs, defining interventions and their outcomes and supporting stakeholder input.

References

View Abstract