Short and long-term costs among women experiencing preterm labour or preterm birth: the German experience

Data source

This study utilised administrative insurance claims data from the Statutory Health Insurance (SHI) sample of AOK Hessen (Versichertenstichprobe AOK Hessen/KV Hessen) [23]. Hessen is a state in central Germany that includes the major cities of Frankfurt and Wiesbaden. The population of the state was estimated at six million individuals in 2012; of these, 1.5 million were insured by AOK. The sample available for research (SHI) is acquired by drawing a random sample of individuals insured by the AOK with a constant selection set of 18.8%. The current SHI sample used in this study included 353,284 persons who were insured in AOK Hessen for at least one day during the five-year period of 2009–2013. The sample is population-based without disease-related selection, with no disease-related dropouts, no recall bias, and a high level of data reliability; this enables patient-based observation and a bottom-up approach to disease costing from the perspective of the health insurance fund. The SHI dataset contains details on healthcare transactions related to insured persons and healthcare providers, including data on care received in general practice, outpatient care (all specialist visits), and hospital care, including emergency visits. Details of this database have been previously published [23, 24].

Study population

We included mothers in the SHI sample who had a recorded diagnosis-related group (DRG) delivery code and a German procedure classification (Operationen- und Prozedurenschlüssel [OPS]) delivery code (Additional file 1) in the relevant study period (1 January 2009—31 December 2013). We further required women to be aged ≥ 12 and < 45 years at delivery and to have at least nine months of medical history available. We excluded women with more than one DRG delivery code within four months, those with no definite date of delivery, and those with a delivery discharge date in 2014. The index date was defined as the delivery date in the eligibility period. Women with multiple pregnancies during the study eligibility period were included in the study cohort once for each delivery, meaning that one woman could appear multiple times within the dataset. The baseline period was defined as the nine months preceding the index date. The delivery hospitalisation for mothers started from the day of hospital admission until the day of hospital discharge. Follow-up started from delivery hospitalisation discharge and lasted until the last date of data collection (31 December 2013), transfer out of the insurance fund, or the death of the mother, whichever occurred first. For women with multiple pregnancies, follow-up after each pregnancy lasted from hospital discharge until the beginning of the next pregnancy (defined as nine months [280 days] before the delivery date of the consequent pregnancy).

It should be noted that not all mothers with PTL delivered a preterm infant, and delivery of a preterm infant was not a condition for being included in the study. This is because not all cases of PTL necessarily result in PTB [2].

Mothers who did not meet the criteria for PTL/PTB were considered non-PTL/PTB mothers. GA was defined using a recorded variable available within the database indicating the expected date of delivery. To calculate the GA, we used this expected date of delivery to calculate an estimated date of conception. This was done by assuming that all pregnancies’ estimated delivery date had been estimated as 280 days after the date of conception. By subtracting 280 days from the expected date of delivery, we derived an estimated date of conception. The difference between this estimated conception date and the actual delivery date was the GA. Additionally, ICD-10 codes present in the mother’s record during birth (P07.2 [extreme immaturity, GA < 28 weeks] and P07.3 [other preterm infants, GA 28–36 weeks] and O09 [duration of pregnancy]) were used to define GA. Mothers with missing GA were assigned to the > 37 weeks of GA based on the distribution of GA in the rest of the population.

Study measures

Statistical methods

Continuous variables were described using average values (median and mean) and measures of data dispersion (interquartile range [IQR], minimum and maximum values). Categorical variables were described using frequencies and percentages. P-values comparing the characteristics of PTL/PTB mothers to non-PTL/PTB mothers were calculated using the chi-squared test/Fisher’s Exact test in those instances when expected cell counts < 5. Univariable negative binomial models were used to estimate the rate of resource utilisation during follow-up as the number of events per person-year with 95% confidence intervals (CI). Costs were presented using summary statistics, which were estimated and included mothers without any resource use (i.e., median and mean costs were estimated, including women who incurred zero costs).

All data programming and analyses were carried out using Microsoft SQL server 2008 (Microsoft, Redmond, WA, USA) and SAS for Windows Release 9.3 (SAS Institute Inc., Cary, NC, USA).

Cases of PTL/PTB

In total, we identified 10,925 mothers in the SHI sample that were eligible for inclusion. An illustration of the cohort selection process can be seen in Fig. 1. Among these 10,925 mothers, 97.7% gave birth to singletons (N = 10,678) and 2.3% to multiples (N = 247). Approximately 19.7% (N = 2147) of the mothers had a diagnosis of PTL during pregnancy or had a PTB. Most mothers (88.2%) had an expected date of delivery variable, which allowed us to directly calculate the GA. A further 1.7% could be categorised into a GA group based on ICD-10 codes, and the remaining 10.1% were assigned to the ≥ 37 weeks GA group as described in the methods section. All mothers who were identified based on giving birth to a low-birth-weight (LBW) infant also had an estimated GA indicating a PTB. Overall, most PTL/PTB mothers delivered at term (37 weeks GA or higher; 65.6%). Approximately 32.3% delivered between 28 and 36 weeks GA, and 2.1% delivered before 28 weeks GA. The proportion of PTL/PTB mothers who delivered at term differed depending on whether the mother gave birth to singletons or multiples (86.4% term for singletons; 41% for multiples).

Maternal characteristics

There were some significant differences between PTL/PTB mothers and non-PTL/PTB mothers (Table 1). Although most mothers were 19–30 years of age when giving birth, the proportion of mothers who were in the youngest (12–18 years of age) or oldest age categories (36–44 years of age) was greater among PTL/PTB mothers compared with non-PTL/PTB mothers (p < 0.0001). PTL/PTB mothers were also more likely to have a CCI score greater than 0 (p = 0.0142), as well as hypertension (p = 0.0002) and diabetes (p = 0.0004). Compared with non-PTL/PTB mothers, those with PTL/PTB were more likely to deliver by Caesarean section (p < 0.0001) and to give birth to twins (p < 0.0001).

Resource use and cost during pregnancy

The median total cost among PTL/PTB mothers during pregnancy was €2130 (mean: €2925). Costs per sector and resource use are presented in Table 2. The highest median costs were observed in the outpatient sector (median: €984; IQR: €752–€1274; mean: €1066), whereas the highest mean costs were observed in the inpatient sector (€1084). As most women were not hospitalised during pregnancy, median costs in the inpatient sector were €0 (IQR: €0–€1483).

Resource use and cost during delivery hospitalisation

During delivery hospitalisation, the mean LOS for PTL/PTB mothers was 6.0 days (median: 4.0; IQR: 3.0–6.0). Mean LOS declined with increasing gestational age, from 13.3 days for < 28 weeks to 8.3 days for 28–36 weeks and 4.7 days for ≥37 weeks (Fig. 2a). Trends were similar when considering median LOS.

Median total costs during delivery hospitalisation were €2037 (mean: €2525.8; IQR: €1614–€3000). Stratified by GA, costs declined with increasing GA; while the cost associated with delivery hospitalisation was €5158 (IQR: €3711–€7488) among mothers delivering with a GA of < 28 weeks, it declined to €2699 for mothers delivering at a GA of 28–36 weeks and €1815 (IQR: €1576–€2536) for mothers giving birth at ≥37 weeks (Fig. 2b). Trends were similar when considering mean costs (Fig. 2b).

Overall, the most common inpatient service used was ‘monitoring and management of a high-risk delivery’ (31%) followed by ‘monitoring and management of a normal birth’ (22%). However, this varied according to the GA. In the < 28 weeks GA group, the most common procedures were Caesarean section (33%) and post-partum instrumental removal of the placenta (16%). In contrast, the most common procedure in the ≥37 weeks group was episiotomy (13%).

Long-term resource use and costs after delivery

The rate of resource use across settings of care in the three years after delivery, split by GA, can be seen in Tables 3, 4 and 5. In the first year after delivery, the largest differences in the rate of resource use was observed for outpatient therapeutic procedures, which declined considerably with increasing GA, from 13.93 (95% CI: 6.22–31.23) uses per person year in the < 28 weeks GA group to 1.13 (95% CI: 0.95–1.35) and 1.07 (95% CI: 0.94–1.22) in the 28–36 weeks and ≥ 37 weeks GA groups, respectively. The rate of use of laboratory tests and basic procedures was also significantly higher in the lowest GA group (Table 3). These differences were less marked during the second and third year after delivery (Table 4 and 5). Although the rate of use of inpatient resources was low, it appeared that the rate of hospitalisations decreased with increasing GA, from 0.29 (95% CI: 0.16–0.53) hospitalisations per person year in the < 28 weeks GA group to 0.15 (95% CI: 0.12–0.20) and 0.13 (95% CI: 0.11–0.15) in the ≥ 37 weeks GA groups, respectively. In contrast to the outpatient resources discussed, these estimates were similar during the second (Table 4) and third (Table 5) years after delivery. However, it should be noted that due to the small number of events, the CIs surrounding these estimates were wide.

Strengths and limitations

This analysis was subject to several limitations. First, as this was a secondary analysis of administrative insurance claims data not collected for research purposes, the availability of some variables was limited. The lack of data on the exact number of physician visits meant that we were restricted to estimating the resource use of ‘at least one physician visit per quarter’ rather than the true number of physician visits. We were also limited by the lack of detailed medical history (such as availability of more specific GAs). To maximise the sample and increase the generalisability of results, we reduced the length of minimum medical history required for inclusion to nine months prior to delivery; this meant we were unable to characterise the obstetric history of women in detail. In addition, we allowed mothers to enter the cohort until the end of the study period, with the trade-off that we would not have long-term follow-up data for women entering toward the end of the study period. While these strategies allowed us to maximise the starting cohort sample size, the sample size decreased with time, as only mothers with recorded deliveries prior to 1 January 2011 (38%) could be potentially observed in the database for the entire three years of follow-up. Although the potential impact of loss to follow-up should be born in mind when interpreting the results, we do not expect significant bias to be introduced, as the main cause of the reduction in the sample size over time was women entering the study close to the end of the study period. The resulting lack of precision and wide CIs, however, is a limitation. A further limitation is the lack of an ability to link data between mothers and infants, which could lead us to miss instances of PTB, and prevented any assessment of combined mother-infant costs. As with any administrative database study, missing or potentially erroneously recorded diagnostic codes is also a limitation.

Our manuscript aimed to assess the total costs associated with all PTL/PTB mothers. However, it is likely that mothers of multiples or mothers with complications during birth have greater resource use and costs compared with mothers of singletons without complications. Further research on how resource use and costs vary not only according to GA, but to the parity and medical history of the mothers, would be of value. In addition, although our aim was to provide a descriptive overview of costs associated with PTL/PTB in Germany, future studies utilising an appropriately constructed control group of non-PTL mothers who deliver at term—to quantify the excess costs associated with PTL/PTB—would also be of interest. We found no published cost estimates of pregnancies leading to term birth in Germany to contrast our estimates of costs associated with PTL/PTB; however, prior comparative studies have found that mothers who experience PTL/PTB have significantly worse health outcomes, [17] longer LOS, [13] and higher resource use [18] compared with mothers who do not experience PTL/PTB, which would likely indicate excess associated costs compared with non-PTL mothers who deliver at term. We were not able to assess the total costs for all PTL/PTB mothers, including costs incurred by their infants, as it was not possible to link all mothers to infants in the AOK database. However, previous studies in the German setting [12] and other countries [8, 36, 37] have found that PTB infants have higher resource use and costs compared with term infants, and that infant costs are generally higher than costs incurred by mothers [13]. This indicates that, had we been able to link all mothers to their infants, the estimated total costs associated with PTL/PTB among mothers and their infants would have been considerably higher. Our analyses were also limited to an assessment of direct costs. It is important to note that PTL/PTB also results in indirect costs, such as earnings lost from taking time off work or costs associated with travel to and from the hospital [30]. It should be noted that costs reflected those incurred within the respective years from 2009 to 2013. Although this may pose a limitation, the consumer price index for healthcare in Germany according to Eurostat [38] during 2009–2013 showed an average annual increase of 1.5%; therefore, the effect of year-on-year changes during our study period was small.

Despite these limitations, our analysis provides up-to-date data on the prevalence and maternal PTL/PTB costs and resource use in a country where limited information on these exist. The apparent trend of declining costs with increasing GA we observed is interesting, and highlights the importance of considering maternal costs, in addition to infant costs, in burden of illness studies of PTL/PTB. Future studies investigating the causes of the increase in costs at lower GAs, particularly over the longer term, would be of great interest. Our analysis is also strengthened by reflecting direct medical costs from the third-party payer perspective, and a long follow-up, relative to previously published studies.