OBJECTIVES: Evidence is mixed regarding the impact of false-positive (FP) newborn bloodspot screening (NBS) results on health care use. Using cystic fibrosis (CF) as an example, we determined the association of FP NBS results with health care use in infants and their mothers in Ontario, Canada.
METHODS: We conducted a population-based cohort study of all infants with FP CF results (N = 1564) and screen-negative matched controls (N = 6256) born between April 2008 and November 2012 using linked health administrative data. Outcomes included maternal and infant physician and emergency visits and inpatient hospitalizations from the infant’s third to 15th month of age. Negative binomial regression tested associations of NBS status with outcomes, adjusting for infant and maternal characteristics.
RESULTS: A greater proportion of infants with FP results had >2 outpatient visits (16.2% vs 13.2%) and >2 hospital admissions (1.5% vs 0.7%) compared with controls; CF-related admissions and emergency department visits were not different from controls. Differences persisted after adjustment, with higher rates of outpatient visits (relative risk 1.39; 95% confidence interval 1.20–1.60) and hospital admissions (relative risk 1.67; 95% confidence interval 1.21–2.31) for FP infants. Stratified models indicated the effect of FP status was greater among those whose primary care provider was a pediatrician. No differences in health care use among mothers were detected.
CONCLUSIONS: Higher use of outpatient services among FP infants may relate to a lengthy confirmatory testing process or follow-up carrier testing. However, increased rates of hospitalization might signal heightened perceptions of vulnerability among healthy infants.
- CF —
- cystic fibrosis
- CIHI —
- Canadian Institute for Health Information
- ED —
- emergency department
- FP —
- NBS —
- newborn bloodspot screening
What’s Known on This Subject:
Infant health care use prompted by false-positive (FP) newborn bloodspot screening results is equivocal and has not been studied in the context of cystic fibrosis. No researchers have examined maternal health care use after infant FP screening results.
What This Study Adds:
Compared with controls, infants with FP cystic fibrosis newborn screening results experience more hospital admissions during a 1-year postpartum period, but their mothers do not experience an increase in health care use.
Newborn bloodspot screening (NBS) involves taking a small blood sample from newborns and testing it for a number of serious conditions that are asymptomatic at birth but require early treatment to prevent significant morbidity.1 Although the US Centers for Disease Control and Prevention hails NBS as 1 of the 10 great public health achievements of the last decade,2 expanded NBS (accompanied by more false-positive [FP] screening results) has heightened a long-standing focus on the patient and system impacts of FP results.3–16
Parental psychosocial response to FP results has been well studied.4,9–16 Findings to date suggest that although distress is likely transient and does not reach levels of clinical concern, the psychosocial experience of receiving FP results is not benign.4,9–16 Some have speculated that plausible psychosocial distress may prompt parents to seek more health care services for their children in response to lingering concerns related to the FP result.4–8 To date, the evidence that reflects on this question is inconsistent, reflects small sample sizes, does not always adequately control for confounding factors related to underlying illness in cystic fibrosis (CF)-positive infants (eg, hypoxia),17 and does not capture all disorders for which screening is available. Using retrospective self-report data on health service use in the metabolic disease context, Waisbren et al4 found increased hospitalizations, whereas Lipstein et al5 found no significant associations with health service use in infants with FP results compared with controls. Findings are similarly mixed from studies in which researchers use administrative claims data; Tarini et al6 found no evidence of increased service use among Medicaid-insured infants with FP results for metabolic, endocrine, or hemoglobin disorders, whereas Karaceper et al7 found increased rates of physician visits and hospitalizations in the first year of life among Canadian infants who received FP results for fatty acid oxidation disorders (ie, medium-chain acyl-coenzyme-A dehydrogenase deficiency) relative to screen-negative controls.
We sought to expand this evidence base while minimizing constraints on sample size by studying CF in a universal-access health care system. In Ontario, CF NBS involves a 2-step process of measuring immunoreactive trypsinogen followed by screening the CF transmembrane regulator gene for 39 mutations. Infants identified as having 1 CF mutation or an immunoreactive trypsinogen >99th percentile are referred for diagnostic sweat-chloride testing and genetic counseling. Of those who resolve to be FP, 80% are CF carriers.18 Our objective was to determine the association of FP NBS results with infant health care use over the first year of life. We also aimed to determine maternal use of health services after the receipt of FP results, an unexplored outcome in the published literature.
This was a population-based, retrospective, matched-cohort study of mother-infant pairs for all infants born between April 1, 2008, and November 30, 2012, in Ontario who had an FP newborn screening test for CF (exposed) or negative results on all other screened conditions (unexposed controls).18,19 One-year outcomes were measured starting at 3 months of age to allow for confirmatory testing. This study was approved by the institutional review boards at Sunnybrook Health Sciences Centre, and The Hospital for Sick Children in Toronto, Canada.
Mother-infant pairs were identified by using the MOMBABY data set, which links newborn birth records to the maternal hospital record on the basis of the Discharge Abstract Database, collected through the Canadian Institute for Health Information (CIHI). NBS data were obtained from the Better Outcomes Registry and Network, Ontario’s pregnancy, birth, and childhood registry and network. Information on inpatient hospitalizations and emergency visits were obtained by using the Discharge Abstract Database and CIHI’s National Ambulatory Care Reporting System, respectively. Outpatient physician visit data were derived from the Ontario Health Insurance Plan physician billing claims. Maternal postal code on the delivery record was linked to 2006 census data to describe rural residence and maternal deprivation quintiles. These data were linked by using unique, encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences by using SAS Enterprise Guide 6.1.
All infants who had (1) FP NBS results for CF and negative results on all other conditions screened in Ontario (exposed) or (2) negative results on all conditions screened (unexposed controls) were eligible. We excluded pairs in which either the infant or mother died before the end of the follow-up period, the infant was hospitalized for longer than 2 months after birth, or matching variables were missing. Infants in the control group were matched 4:1 with infants in the exposure group on the basis of the infants’ year of birth, sex, rural residence, and material deprivation quintile. For each patient in the exposure group, a greedy matching algorithm was used to select the control that most closely matched that patient in terms of the 4 matching factors. Controls were randomly selected if >4 possible controls were identified.
Because the time from an NBS test to a confirmatory test typically takes 2 months in Ontario, our follow-up period started 3 months after birth. Ninety-five percent of FP cases are resolved as such by this point in time (J. Milburn, MHA, personal communication, 2017). Infant outcomes included infant outpatient physician visits (routine well-infant immunization visits, other primary care visits, and all other outpatient visits), emergency department (ED) visits (overall, low acuity [Canadian Triage and Acuity Scale levels 4 and 5] and high acuity [Canadian Triage and Acuity Scale levels 1–3]), and hospital admissions within 1 year of the start of follow-up. Maternal outcomes included outpatient physician visits (routine postpartum and mental health visits per validated definition20; Supplemental Table 5) and mental health ED visits within 1 year. For maternal outcomes, we conducted a sensitivity analysis in which follow-up started 30 days after delivery.
Covariates included maternal age and gestational age at delivery, multifetal pregnancy, major congenital anomaly, length of infant’s postnatal hospital stay, maternal history of mental health problems (hospitalization, ED visits, or outpatient visits) in the 3 years before delivery (Supplemental Table 5), rural residence, material deprivation quintile, primary care provider type at 3 months of age, and maternal age at first birth (a more precise risk factor for poor infant outcomes than age at the index pregnancy).21 Major congenital anomalies were defined by using CIHI case mix groups (Supplemental Table 6). Length of hospital stay categories were defined clinically (ie, <1 week representing uncomplicated births, including by cesarean section, 1–2 weeks, >2 weeks). Material deprivation is a component of the Ontario Marginalization Index,22 a geographically based index developed to measure various aspects of socioeconomic status at the level of a dissemination area (400–700 inhabitants). Primary care providers (including family physicians and pediatricians) were identified as in previous work by using physician rostering information as well as data on frequency of primary care visits.23
Baseline characteristics and outcomes of mother-infant pairs were compared according to screening test results (FP or screen-negative). We tested differences in maternal and infant outcome frequencies using χ2 tests. Associations between an FP CF NBS test and the outcomes were examined with conditional negative binomial regression models. Infant models included gestational age, length of hospital stay, month of birth, multifetal pregnancy, major congenital anomaly, material deprivation quintile, and rural residence as covariates. Maternal models included maternal history of mental health problems, material deprivation quintile, and rural residence as covariates. We hypothesized that the relationship between CF NBS test results and infant health care use may depend on whether the primary care provider is a family physician or a pediatrician, so we also tested this interaction in the models and subsequently ran stratified models in which P values for the interaction terms were significant at <.05.
We identified 625 417 births during the study period, of which 606 650 (97.0%) had an NBS record. Of these mother-infant pairs, 12 162 were not eligible for health care coverage; in 283 cases, the infant (n = 149) or mother (n = 134) died before the follow-up period; 860 infants had a hospital stay of >2 months; 8469 were missing information on matching variables; and 198 infants were positive for other screening tests, leaving 584 678 (96.4%). Of these 584 678, 1564 (0.27%) children had FP NBS results for CF during the study period, and 6256 infants were selected as the screen-negative control cohort. Baseline characteristics were similar across FP and screen-negative groups (Table 1). The vast majority of children in both groups were singletons, born at ≥37 weeks’ gestation with no congenital anomalies, had short hospital stays, and were cared for by family physicians. The majority of mothers were >20 years of age at first birth and approximately half of the mothers in each group had received outpatient mental health visits in the 3 years before the index birth.
Health Service Use Among Infants
Overall, the majority of infants with FP and screen-negative results had 2 or more well-infant immunization visits over the 1-year follow-up period (Table 2). Similarly, majorities in both groups had at least 2 nonimmunization-related primary care visits (Table 2).
Although majorities in both groups had fewer than 2 specialist visits over the course of the follow-up period (Table 2), infants with FP results were more likely to have visited a specialist. When we examined these visits according to specialty, consultant pediatrician visits constituted the greatest percentage in both groups (FP: 37%; screen-negative: 40%). Infants who had FP results were more likely than infants with screen-negative results to have visited a medical geneticist (1.8% vs 0.7%) or a pediatric respirologist (1.0% vs 0.6%).
Of infants, ∼40% visited the ED at least once in the outcome year; however, there were no differences between groups with respect to frequency of ED visits (Table 2). Frequencies of ED visits were similar in the distribution of low- and high-acuity visits.
Although only ∼5% of infants were admitted to hospitals in the study period, FP infants were more likely to have had an admission (5.5% vs 4.2%) (Table 2). Indications for admission were not significantly different between the groups. Of the top 5 admission diagnoses, 4 were the same between groups (ie, bronchiolitis, urinary tract infections, gastroenteritis, and pneumonia), but for FP infants, failure to thrive was among the top 5.
Health Service Use Among Mothers
Of mothers in both groups, ∼74% had a routine postpartum or primary care visit (Table 3). However, 398 (25.5%) mothers of infants with an FP result had mental health visits compared with 1347 (21.5%) mothers of infants with screen-negative results. There were no differences in mental health-related ED visits. When maternal outcomes were measured beginning at 30 days after delivery in our sensitivity analysis, the findings were the same.
Adjusted Models: Infants
After adjusting for potential confounders in the regression analyses for infants, we found that FP status was independently associated with higher rates of specialist visits (relative risk 1.4; 95% confidence interval 1.22–1.62) and infant hospital admissions (relative risk 1.67; 95% confidence interval 1.21–2.31) (Table 4). FP status was not associated with rates of well-infant immunization visits, other primary care visits, ED visits, or maternal mental health-related visits.
For specialist visits, ED visits, and hospital admissions, we identified a significant interaction between FP status and primary care provider. Stratified models indicated that the effect of FP status was greater among those whose primary care provider is a pediatrician than those cared for by a family physician (Table 4).
Adjusted Models: Mothers
After adjusting for all other predictors of service use in the regression analyses for mothers, we found infant FP status was not associated with an increased rate of maternal outpatient mental health visits or maternal mental health and ED visits combined. The results of the sensitivity analysis with follow-up starting 30 days after delivery were similar (Table 4).
In this large, population-based study of both infant and maternal health care use after FP newborn screening results, we show (after adjustment for potential confounders) that infants with FP newborn screening results for CF have a higher frequency of specialist visits and hospitalizations, but not ED visits, as compared with screen-negative controls during a 1-year follow-up period. We also show that mothers of infants with FP results did not have more primary care postpartum visits, outpatient mental health visits, or mental health–related ED visits during a 1-year postpartum period compared with mothers of screen-negative infants.
Our findings align most closely with Karaceper et al’s7 Ontario-based study of infants who received FP newborn screening results for medium-chain acyl-coenzyme-A dehydrogenase deficiency. Using administrative records, the authors found that compared with screen-negative controls, FP infants (n = 43) experienced significantly higher rates of physician visits and hospitalizations in their first year of life. As with our study, the samples were population based and in the same universal-access health system. Our findings do not align with Tarini et al’s6 finding of similar health service use patterns among infants with FP and screen-negative metabolic, endocrine, or hemoglobin disorder–related screening results. However, that study included only infants insured through Medicaid in Michigan and so may be less representative of a more socioeconomically diverse sample. Similarly negative findings by Lipstein et al5 were also from a predominantly low-income US population in which both real and perceived barriers to care may result in less use compared with a universally accessible system reflected herein. Our findings are consistent with Waisbren et al,4 who found increased hospitalizations among children with FP newborn screening results compared with screen-negative controls. To our knowledge, this is the first study to report on health service use among infants with FP results for CF, the largest single-disease contributor to newborn screen-positive results18 and the first study to report on health service use among mothers of FP infants.
Although studies in which researchers assess parental psychosocial response to FP newborn screening results suggest that distress is likely transient and does not reach levels of clinical concern, there are both quantitative and qualitative signals from this literature indicating that this experience is not emotionally benign for families.4,9–16 Health-seeking behavior among those with FP results may be explained by CF-carrier testing for family members because 80% of infants with FP CF screening results are carriers.12,18 Alternatively, these visits may represent care that is related to confirmatory testing and diagnostic follow-up, particularly for families who receive complex genotypes that require further investigation or live in remote areas, for whom it may take longer than 3 months to complete the NBS protocol. Finally, it is possible that perceptions of uncertainty, worry about the long-term, or perceived infant vulnerability could explain increased use of respirology and genetics services.9
However, these factors do not explain the increased rate of hospital admissions among FP infants. Because our follow-up window did not start until 3 months after the report of the screening result to allow for the confirmatory testing processes to take place for the majority, admissions are unlikely to be related to ongoing diagnostic testing. To our surprise, the effect of FP status on specialist, ED, and hospital visits was greater among those cared for by pediatricians than those cared for by family physicians. Our speculation that family physicians, possibly having less exposure to CF, might be quicker to refer to tertiary care24 is not supported by these findings. As suggested by Karaceper et al,7 it is possible that residual confounding related to underlying illness among FP infants or physiologic differences between the groups may explain the increased rate of hospital admissions observed in our cohort. Or, possibly, our findings may reflect different health care–seeking behavior among parents who chose pediatricians for primary care. It may also reflect physician (mis)understanding of the health risks associated with CF FP status despite the fact that CF is better known to primary care providers than many of the other, rarer NBS targets. Finally, results from these studies may also reflect the limitations in the clinical and psychosocial details available in administrative records.
Although ongoing efforts are required to support parents through the NBS journey, and system design is central to their experience,9 we are reassured by the finding that no elevated use of primary care, mental health, or ED services was detected among mothers of FP infants. Because this is the first study in which researchers explore maternal health care use in the context of FP NBS results, and because the related evidence is equivocal, further research is warranted to generate a finer-grained understanding of the role of underlying health status provider practice patterns and parental psychosocial response in downstream health service use after newborn screening. If increased use is in fact identified in other cohorts and revealed to be a predictable consequence of expanded newborn screening, efforts to mitigate unnecessary use should be invoked, and downstream system costs should be considered in NBS costing analyses.
Compared with controls, we found that infants who have FP CF NBS results experience more hospital admissions during a 1-year postpartum period, but their mothers do not experience a similar increase in health care use. Possible explanations for the infant admission findings relate to parental psychosocial response, physician perceptions and practice patterns, and residual confounding. Optimal delivery of a complex health intervention such as NBS requires a system of care approach to evaluation,24–26 one that accounts for patient, provider, and system impact. By understanding the downstream experience for patients, providers, and the health care system, newborn screening programs can better support patients and providers in navigating screening results. Specifically, counseling and education resources can be improved to ensure that parents and providers understand that FP CF results (ie, including carrier status) are medically benign for the index case but have legitimate implications for cascade family testing and reproductive planning. When further support from the health care system continues to be sought by families who receive these results, newborn screening programs ought to consider the impact of related costs incurred and saved as part of routine program evaluation. As additional disorders meet the criteria for screening panels, challenges imposed by FP results will continue to warrant attention to maximize the benefits and minimize the burdens associated with this important public health program.
- Accepted August 11, 2017.
- Address correspondence to Robin Z. Hayeems, ScM, PhD, Hospital for Sick Children Research Institute, 686 Bay St, Room 11.9710, Toronto, ON M5G 0A4. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by the Canadian Institutes of Health Research (operating grant 106505). Dr Guttmann is supported in part by a Canadian Institutes of Health Research Applied Chair in Reproductive and Child Health Services and Policy Research. This study was supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. The opinions, results, and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by the Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred. Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information. However, the analyses, conclusions, opinions, and statements expressed herein are those of the authors and not necessarily those of the Canadian Institute for Health Information.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Grosse SD,
- Boyle CA,
- Kenneson A,
- Khoury MJ,
- Wilfond BS
- Tarini BA,
- Clark SJ,
- Pilli S, et al
- Utah Department of Health
- Newborn Screening Ontario
- Newborn Screening Ontario
- Cohen E,
- Gandhi S,
- Toulany A, et al
- Hayeems RZ,
- Miller FA,
- Carroll JC, et al
- Raffle AE,
- Gray JAM
- Hayeems RZ,
- Chakraborty P
- Copyright © 2017 by the American Academy of Pediatrics