Quality in perinatal care: applying performance measurement using joint commission on accreditation of healthcare organizations indicators in Italy

Maternal and child health is a public health priority, because pregnancy, childbirth and puerperium are leading causes of hospitalization for women, and birth-related events are internationally considered to be among the best measures for assessing health-care quality. The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) developed the Perinatal Care (PC) core measure set that includes five metrics with sufficient evidence that better performance are clinically important and are possible with system and process improvement [1, 2]. These core measures were chosen from a broader set among those recommended by the National Quality Forum (NQF) by a technical advisory panel of experts in perinatal care. The benefit of the core measures is that they provide a national, standardized set of quality metrics that hospitals can use [3].

In Italy, since 2010, the Outcomes National Plan (Piano Nazionale Esiti - PNE), started up by the National Agency for Regional Health Services (Agenzia Nazionale per i Servizi Sanitari Regionali-Age.Na.S.), has provided an active evaluation of hospital performance, but the PNE’s indicators regarding perinatal care focus attention on cesarean section only [4]. Also, the 2011–2013 National Health Care Plan underlines the need for development and implementation of certification programs for hospital birth centers, by involving scientific societies, as well as associations of obstetricians and nurses [5, 6].

The quality of care provided to the adult hospitalized Italian population has been scrutinized in the past years by use of adequate indicators [7, 8], whereas only sparse data is available for perinatal care, although an interesting approach suggesting the use of a set of 19 indicators for the performance assessment in the maternity pathway in one region of Italy has thoroughly taken into account also perinatal care indicators [9]. In this context, the primary aim of this study was to assess the quality of perinatal care in a specific geographical area of Italy using the JCAHO indicators, since, compared with PNE indicators, they involve more aspects of perinatal care and include indicators for which there is no standardized monitoring system in Italy, such as breastfeeding or elective delivery [10]. Further aims of the study were to assess the feasibility of these quality indicators in our healthcare setting that could be used to monitor the effects of quality improvement interventions, and to analyze whether maternal, pregnancy care and neonatal characteristics could be associated with the quality of perinatal care hospital performance, measured through JCAHO PC indicators.

The study was carried out by retrospectively reviewing medical charts of women over the age of 18 who experienced delivery in Gynecology/obstetrics wards between January 1 and December 31 2016, and those of their newborns hospitalized in the Neonatology or Neonatal Intensive Care Unit (NICU) of a public non-teaching hospital in Catanzaro (Italy).

The medical charts of the women were matched with those of their newborns and reviewed concurrently. Medical charts were selected according to the list of the International Classification of Diseases, 10th Revision, Procedure Coding System (ICD-10-CM/PCS) codes and other diagnosis and procedures codes.

We used the JCAHO perinatal core measure set, which considers the following: Elective deliveries (PC-01); Cesarian section (PC-02) in nulliparous women with a term, singleton baby in a vertex position (NTSV); Antenatal steroids (PC-03); Healthcare-associated bloodstream infections in newborns (PC-04) and Exclusive breast milk feeding (PC-05).

Indicators were calculated according to the methodology specified in the manual for JCAHO measures and summarized in Additional file 1. Whenever the condition described by one of the indicators appeared in the medical record, a score of 1 was assigned if the procedure had been performed consistently with that defined by the indicator, otherwise a score of 0 was attributed.

Since the manual for JCAHO measures does not provide clear target rates, we have used the updated reference goals proposed by NQF and Healthy People 2020, to compare calculated indicators. Regarding the PC-01 indicator, although the optimal rate is considered to be 0% [11], Clark et al. considered that the rate of this indicator will never be and should never be consistently zero [12]. For the PC-02 indicator Healthy People 2020 chose a target of a 23.9% NTSV rate [13]. For the PC-03 and PC-04 indicators, the optimal rate is considered the best performance, that is 100 and 0%, respectively [14, 15]. Finally, for PC-05, a goal of 75% is considered acceptable [16].

Two physicians not involved in patient’ care, but who had been acquainted with the specification manual released at the time of the study design, collected the data and retrieved them on a standardized electronic report form. Maternal, obstetrical and neonatal data were obtained. Specifically, information included socio-demographic and clinical characteristics, obstetric history and pregnancy, delivery, and characteristics of the newborn.

A total of 1943 women records were reviewed and, of these, 1172 were eligible for at least one of the PC-Mothers measures (PC-01, 02 and 03). Considering a total of 36 twin-births of which 5 with a single eligible newborn, 1974 newborns’ records were identified and reviewed in order to be included in one of the Newborn PC (PC-04 and 05) subpopulations. The medical records of 297 women were reviewed for the “Elective delivery” indicator, 904 for the “Cesarean section” indicator and 31 for the “Antenatal steroids” indicator. Moreover, 473 newborns’ medical records were reviewed for the “Healthcare-associated bloodstream infections in newborns” indicator, and 1687 for the “Exclusive breast milk feeding” indicator.

Frequency distribution of maternal, pregnancy and prenatal care characteristics according to “Elective delivery”, “Cesarean section” and “Exclusive breast milk feeding” rates are shown in Additional file 2, whereas the scoring of the five PC indicators is reported in Table 1.

Perinatal care quality and safety is a complex process aimed at achieving maximum health potential for the fetus, the newborn and the mother, and, similarly, evaluating quality of perinatal care is complex because it involves different populations. This is the first study performed in Italy that has evaluated the quality of perinatal care by using all the five JCAHO PC indicators. The application of this feasible set of indicators allowed us to measure several aspects of perinatal care for which there is no standardized monitoring system in Italy, as well as the factors associated with an eventual suboptimal performance. Overall, the results pointed out that the quality of perinatal hospital care, measured through the JCAHO PC indicators, is indicator dependent, with exclusive breastfeeding performing well, whereas for most indicators there is room for improvement.

The most critical result of our study pertains to the elective/early-term delivery, that has been performed in 27.6% of eligible women, far from the goal (0%) set by Clark et al., that strongly confirmed the commitment to the elimination of early term elective delivery [12]. This result is concerning, since this indicator has been reported to be one of the most important performance measures due to its impact on clinical practice, on healthcare costs, and on patients’ morbidity [17]. Albeit, Salemi et al. in a recent cohort study, highlighted that no excess risk of respiratory morbidities, neonatal sepsis, and NICU admission was associated with the elective induction of delivery at 37–38 weeks of gestation with respect to infants expectantly managed and delivered at 39–40 weeks. Only the early cesarean section significantly supports 13 to 66% increase of several adverse outcomes occurrence in neonates, when compared with the full-term group [18].

In Italy, the practice of induction of labor and elective cesarean section are investigated separately. The Europeristat project has detected that inductions were performed in 15.9% of the total births in Italy in 2010, but data do not allow a separate evaluation of the induction performed in early term deliveries [19]. Therefore, there is a need to standardize definitions and evaluation methods of elective early term deliveries, in order to improve the validity of comparisons among countries. Moreover, as reported by Clark et al. [17], the calculation of this indicator is prone to errors particularly related to the selection of clinical indications for inclusion/exclusion of elective delivery.

The result of the PC-02 indicator is in line with the Italian national figure of cesarean section rate that is recorded as one of the highest among European countries [20]. Several reasons may be related to the high cesarean rate found in our study: first of all, the cesarean section high rate may be associated with the changed attitude of the physicians aimed at reducing exposure to malpractice litigation [21]. Indeed, in our results cesarean delivery was significantly more frequent in pregnancy with AROM, meconium-stained amniotic fluid and newborns with > 4000 g birth weight, that are not medical indications for cesarean section, suggesting therefore that choice to perform cesarean section was related to a cautious approach of physicians to delivery. Moreover, women’s choice of cesarean delivery is increasing since it is perceived as an effective procedure to avoid pain and the other disadvantages associated with vaginal delivery [22, 23]. Finally, cesarean section has been reported to be more frequently performed in the private healthcare sector than in the public one [24‐26] and one of the main reasons for this is that cesarean sections receive higher reimbursement than normal vaginal births, regardless of the risks to women [27] and private healthcare facilities are commonly involved in deliveries in Southern Italy [24, 28, 29].

It is well known that cesarean sections should be discouraged because they create serious complications for mothers such as infections [30], obstetric haemorrhage [31], uterine rupture, stillbirth and pre-term birth [32]. Also, there is emerging evidence that children born by cesarean section have an increased risk of altered immune development, allergy and asthma, and reduced intestinal gut microbiome diversity [32]. Moreover, as our results have shown and according to previous studies, cesarean sections have a negative effect on the exclusive breast milk feeding, probably because they limit the practice of rooming-in, so delaying the mother-child interaction [33, 34]. Thus, cesarean section should not be considered an alternative to vaginal delivery, and should be viewed with caution. Indeed, among the measures taken to discourage unnecessary cesarean sections, several countries have also narrowed the gap in hospital payment between a cesarean section and a vaginal birth [35].

The exclusive breast milk feeding indicator achieved the best performance (81%) in our study. In particular, our result is higher than the set goal (75%) [16], and higher than the value found in 2015 in Italy (77%) [36]. However, it is well known that in Italy there is a tendency to wean children from breastfeeding at an early age, on average at 4 months [29] although, as previously reported, the start of exclusive breast milk feeding during hospitalization positively influences the continuation of this practice in the following months [33, 34, 37].

Health professionals’ approaches to breastfeeding during antenatal care are crucial to promote exclusive breast milk feeding. Our results highlight that exclusive breast milk feeding significantly increased in women who underwent prenatal tests; probably this practice contributed in encouraging increasing educational activities in physician-patient relationship also regarding breastfeeding. Also, avoiding in-hospital formula supplementation appears to be a key step for breastfeeding success, together with the appropriate implementation of the other Baby Friendly Hospital Initiative (BFHI) steps [38]. As highlighted by recent meta-analyses, the BFHI approach by steps requiring implementation at the maternity ward, followed by home and family support through counselling, appears to be crucial for breastfeeding success in expectant and/or nursing mothers [39, 40]. The significantly lower adherence to PC-05 of newborns with a birth weight < 2500 g as well as in women with previous cesarean section, AROM and, as previous mentioned, with cesarean delivery, are important concerns of our results. These findings suggest the involvement of several non-clinical factors that would seem to be attributable to the overly cautious attitude of the physician concerning patients’ management, suggesting the need for improvement in the training of healthcare professionals.

Antenatal steroids are intended to reduce the burden of prematurity-related illness (respiratory distress, intraventricular haemorrhage, necrotizing enterocolitis, and patent ductus arteriosus) in preterm newborns. The prevalence of early preterm infants revealed by our study (1.8%) is in line with previous data [41], nevertheless there are concerns in international comparisons for this quality measure due to the differences in registration criteria and definitions across countries [42]. This indicator was satisfied in only 87.1% of the eligible patients, highlighting a suboptimal process of care that might have led to increased morbidity or mortality. Indeed, in our study, when the PC-03 indicator was not satisfied, newborns more frequently experienced mechanical ventilation compared with newborns whose mother received antenatal steroids; however, these results are to be interpreted with caution, because of the limited number of included patients.

As underlined in a recently published meta-analysis, there is continuing uncertainty about the most appropriate method to calculate the healthcare-associated bloodstream infections burden in NICUs. Cumulative incidence of these severe complications is reported to be variable from 2.9 to 22.8% [43]. Only two healthcare-associated bloodstream infections occurred during 2016 (0.4%) and this result allows us to interpret compliance with the PC-04 indicator as to be satisfied, although also in this case a cautious interpretation of the results is needed regarding the above mentioned concerns in the calculation of this indicator.

Although the application of JCAHO PC quality indicators was feasible and intuitive, results of this study should be evaluated in light of potential limitations, considering the fragmentary availability of required data. First, the poor comparability among multiple classification systems is the most substantial barrier that we met. Gilbert et al. suggested that improvement of performance’s quality depends on the improvement of the accuracy of data recording and its transparency [1]. Second, patients were recruited from a hospital located in Southern Italy, and may not be representative of the entire country. Third, most of the previous studies were conducted on large numbers of hospitals and therefore were based on aggregated data. Instead, by focusing on one hospital, our results were derived from a smaller number of patients, but detailed information was gathered from each of them.

In conclusion, our findings revealed significant deficiencies in the adherence to recommended processes of perinatal care and, consistently with previous studies conducted by some of us to estimate the adherence to evidence-based processes of care in several settings [7, 44, 45], suggest that it is essential to increase efforts to implement evaluation processes that reflect the healthcare quality based on current evidence and related practice guidelines. The application of the JCAHO PC indicators has demonstrated to be feasible, intuitive and useful to measure perinatal hospital performance, and, although the poor comparability among multiple available quality measures represents a barrier, these performance metrics can be reliably used within an institution, thus enabling comparisons of performance over time, particularly after the implementation of quality improvement interventions.