Accuracy of international classification of diseases, ninth revision, codes for postpartum hemorrhage among women undergoing cesarean delivery

The Association of Antidepressants in Late Pregnancy with Postpartum Hemorrhage: Systematic Review of Controlled Observational Studies

Introduction: Despite advances in obstetric care, postpartum hemorrhage (PPH) is a leading cause of maternal mortality worldwide. Prior reviews of studies published through 2016 suggest an association of antidepressant use during late pregnancy and increased risk of PPH. However, a causal link between prenatal antidepressants and PPH remains controversial. Objectives: This systematic literature review aimed to synthesize the empirical evidence on the association of antidepressant exposure in late pregnancy with the risk of PPH, including studies published before and after 2016. Methods: A systematic literature search was conducted using PubMed, OVID Medline, EMBASE, SCOPUS, PsycINFO, and CINAHL from inception to September 9, 2023. Original, peer-reviewed studies (published in English) that reported on the frequency or risk of PPH in women with evidence of antidepressant use during pregnancy and included at least one control group were included. Results: Twenty studies (eight published after 2016) met inclusion criteria, most of which focused on the risks of PPH associated with selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepinephrine reuptake inhibitors (SNRIs). The main findings from the individual studies were mixed, but the majority documented statistically significant associations of PPH with late prenatal exposure, especially for exposures occurring within 30 days of delivery, compared with unexposed deliveries. Fourteen studies addressed underlying antidepressant indications or their correlates. Few studies focused on prenatal antidepressants and the risk of well-defined severe PPH or on antidepressant dose changes and general PPH risk. None examined competing risks of antidepressant discontinuation on mental health outcomes. Conclusions: Late pregnancy exposure to antidepressants may be a minor risk factor for PPH, but it is unclear to what extent reported associations are causal in nature, as opposed to correlational (effects related to nonpharmacological factors including maternal indication). For patients needing antidepressants during pregnancy, current evidence does not favor routinely discontinuing antidepressants specifically to reduce the risk of PPH.

Zero-shot interpretable phenotyping of postpartum hemorrhage using large language models

Many areas of medicine would benefit from deeper, more accurate phenotyping, but there are limited approaches for phenotyping using clinical notes without substantial annotated data. Large language models (LLMs) have demonstrated immense potential to adapt to novel tasks with no additional training by specifying task-specific instructions. Here we report the performance of a publicly available LLM, Flan-T5, in phenotyping patients with postpartum hemorrhage (PPH) using discharge notes from electronic health records (n = 271,081). The language model achieves strong performance in extracting 24 granular concepts associated with PPH. Identifying these granular concepts accurately allows the development of interpretable, complex phenotypes and subtypes. The Flan-T5 model achieves high fidelity in phenotyping PPH (positive predictive value of 0.95), identifying 47% more patients with this complication compared to the current standard of using claims codes. This LLM pipeline can be used reliably for subtyping PPH and outperforms a claims-based approach on the three most common PPH subtypes associated with uterine atony, abnormal placentation, and obstetric trauma. The advantage of this approach to subtyping is its interpretability, as each concept contributing to the subtype determination can be evaluated. Moreover, as definitions may change over time due to new guidelines, using granular concepts to create complex phenotypes enables prompt and efficient updating of the algorithm. Using this language modelling approach enables rapid phenotyping without the need for any manually annotated training data across multiple clinical use cases.

Placenta Accreta Spectrum

Placenta accreta spectrum (PAS) is one of the most dangerous conditions in pregnancy and is increasing in frequency. The risk of life-threatening bleeding is present throughout pregnancy but is particularly high at the time of delivery. Although the exact cause is unknown, the result is clear: Severe PAS distorts the uterus and surrounding anatomy and transforms the pelvis into an extremely high-flow vascular state. Screening for risk factors and assessing placental location by antenatal ultrasonography are essential for timely diagnosis. Further evaluation and confirmation of PAS are best performed in referral centers with expertise in antenatal imaging and surgical management of PAS. In the United States, cesarean hysterectomy with the placenta left in situ after delivery of the fetus is the most common treatment for PAS, but even in experienced referral centers, this treatment is often morbid, resulting in prolonged surgery, intraoperative injury to the urinary tract, blood transfusion, and admission to the intensive care unit. Postsurgical complications include high rates of posttraumatic stress disorder, pelvic pain, decreased quality of life, and depression. Team-based, patient-centered, evidence-based care from diagnosis to full recovery is needed to optimally manage this potentially deadly disorder. In a field that has relied mainly on expert opinion, more research is needed to explore alternative treatments and adjunctive surgical approaches to reduce blood loss and postoperative complications.

Zero-shot Interpretable Phenotyping of Postpartum Hemorrhage Using Large Language Models

Many areas of medicine would benefit from deeper, more accurate phenotyping, but there are limited approaches for phenotyping using clinical notes without substantial annotated data. Large language models (LLMs) have demonstrated immense potential to adapt to novel tasks with no additional training by specifying task-specific i nstructions. We investigated the per-formance of a publicly available LLM, Flan-T5, in phenotyping patients with postpartum hemorrhage (PPH) using discharge notes from electronic health records ( n =271,081). The language model achieved strong performance in extracting 24 granular concepts associated with PPH. Identifying these granular concepts accurately allowed the development of inter-pretable, complex phenotypes and subtypes. The Flan-T5 model achieved high fidelity in phenotyping PPH (positive predictive value of 0.95), identifying 47% more patients with this complication compared to the current standard of using claims codes. This LLM pipeline can be used reliably for subtyping PPH and outperformed a claims-based approach on the three most common PPH subtypes associated with uterine atony, abnormal placentation, and obstetric trauma. The advantage of this approach to subtyping is its interpretability, as each concept contributing to the subtype determination can be evaluated. Moreover, as definitions may change over time due to new guidelines, using granular concepts to create complex phenotypes enables prompt and efficient updating of the algorithm. Using this lan-guage modelling approach enables rapid phenotyping without the need for any manually annotated training data across multiple clinical use cases.

Utilizing International Classification of Diseases Codes to Identify Shoulder Dystocia and Neonatal Brachial Plexus Injury

BACKGROUND

The utilization of International Classification of Diseases, Ninth or Tenth Revision, (ICD-9/10) coding to identify the incidence of disease is frequently performed in medical research. This study attempts to assess the validity of using ICD-9/10 codes to identify patients with shoulder dystocia (SD) with concurrent neonatal brachial plexus palsy (NBPP).

METHODS

This retrospective cohort study examined patients evaluated at the University of Michigan Brachial Plexus and Peripheral Nerve Program (UM-BP/PN) from 2004 to 2018. We reported the percentage of patients with reported NBPP ICD-9/10 and SD ICD-9/10 discharged at birth who were later diagnosed with NBPP by a specialty clinic by interdisciplinary faculty and staff utilizing physical evaluations and ancillary testing such as such as electrodiagnostics and imaging. The relationship of reported NBPP ICD-9/10, SD ICD-9/10, extent of NBPP nerve involvement, and NBPP persistence at age two years were examined via chi-square or Fischer exact test.

RESULTS

Of the 51 mother-infant dyads with complete birth discharge records evaluated at the UM-BP/PN, 26 (51%) were discharged without an ICD-9/10 code documenting NBPP; of these 26 patients, only four had ICD-9/10 documentation of SD at discharge, which left 22 patients with no ICD-9/10 code documentation of either SD or NBPP (43%). Patients with pan-plexopathy were more likely to be discharged with an NBBP ICD-9/10 code than those infants with upper nerve involvement (77% vs 39%, P < 0.02).

CONCLUSION

Use of ICD-9/10 codes for the identification of NBPP appears to undercount the true incidence. This underestimation is more pronounced for milder forms of NBPP.

Hospital-level variation in rates of postpartum hemorrhage in California

BACKGROUND

To examine the extent of hospital-level variation in risk-adjusted rates of postpartum hemorrhage (PPH).

STUDY DESIGN AND METHODS

We performed a cross-sectional study examining live births in 257 California hospitals between 2011 and 2015 using linked birth certificate and maternal discharge data. PPH was measured using International Classification of Diseases Codes version 9. Mixed-effects logistic regression models were used to examine the presence and extent of hospital-level variation in PPH before and after adjustment for patient-level risk factors and select hospital characteristics (teaching status and annual delivery volume). Risk-adjusted rates of PPH were estimated for each hospital. The extent of hospital variation was evaluated using the median odds ratio (MOR) and intraclass correlation coefficient (ICC).

RESULTS

Our study cohort comprised 1,904,479 women who had a live birth delivery hospitalization at 247 hospitals. The median, lowest, and highest hospital-specific rates of PPH were 3.48%, 0.54%, and 12.0%, respectively. Similar rates were observed after adjustment for patient and hospital factors (3.44%, 0.60%, and 11.48%). After adjustment, the proportion of the total variation in PPH rates attributable to the hospital was low, with a MOR of 2.02 (95% confidence interval [CI]: 1.89-2.15) and ICC of 14.3% (95% CI: 11.9%-16.3%).

DISCUSSION

Wide variability exists in the rate of PPH across hospitals in California, not attributable to patient factors, hospital teaching status, and hospital annual delivery volume. Determining whether differences in hospital quality of care explain the unaccounted-for variation in hospital-level PPH rates should be a public health priority.

Mild thrombocytopenia and the risk for postpartum hemorrhage in twin pregnancies

OBJECTIVE

We aimed to investigate the association of mild thrombocytopenia with postpartum hemorrhage (PPH) and blood transfusion among women with twin gestations.

METHODS

A retrospective cohort study (Jan 2015 to May 2019) was performed. Women with twin pregnancies and pre-delivery mild thrombocytopenia were compared to those with normal platelet count. The primary outcome was the rate of PPH, defined as a composite of one or more of the following: (1) need for packed red blood cell transfusion; (2) postpartum hemoglobin decline of ≥3 g/dL; and (3) the use of postpartum uterotonics agents in addition to oxytocin.

RESULTS

Of 1085 women who were included in final analysis, 315 (30.9%) had mild thrombocytopenia (and 770 (69.1%) served as controls. The rate of PPH was increased in the study group (14% vs. 9.4%, P = 0.03), as was the use of uterotonic agents (3.8% vs. 1.3%, respectively, P = 0.02). The rate of blood product transfusion and hemoglobin decline >3 g/dL was not significantly different between the groups. In multivariate logistic regression analysis, mild thrombocytopenia was associated with a higher risk for PPH (OR 1.55 [95% CI 1.02-2.35], P = 0.02).

CONCLUSION

Mild thrombocytopenia in twin pregnancies is associated with an increased risk of interventions such as the use of uterotonic agents.

A comprehensive digital phenotype for postpartum hemorrhage

OBJECTIVE

We aimed to establish a comprehensive digital phenotype for postpartum hemorrhage (PPH). Current guidelines rely primarily on estimates of blood loss, which can be inaccurate and biased and ignore complementary information readily available in electronic medical records (EMR). Inaccurate and incomplete phenotyping contributes to ongoing challenges in tracking PPH outcomes, developing more accurate risk assessments, and identifying novel interventions.

MATERIALS AND METHODS

We constructed a cohort of 71 944 deliveries from the Mount Sinai Health System. Estimates of postpartum blood loss, shifts in hematocrit, administration of uterotonics, surgical interventions, and diagnostic codes were combined to identify PPH, retrospectively. Clinical features were extracted from EMRs and mapped to common data models for maximum interoperability across hospitals. Blinded chart review was done by a physician on a subset of PPH and non-PPH patients and performance was compared to alternate PPH phenotypes. PPH was defined as clinical diagnosis of postpartum hemorrhage documented in the patient's chart upon chart review.

RESULTS

We identified 6639 PPH deliveries (9% prevalence) using our phenotype-more than 3 times as many as using blood loss alone (N = 1,747), supporting the need to incorporate other diagnostic and intervention data. Chart review revealed our phenotype had 89% accuracy and an F1-score of 0.92. Alternate phenotypes were less accurate, including a common blood loss-based definition (67%) and a previously published digital phenotype (74%).

CONCLUSION

We have developed a scalable, accurate, and valid digital phenotype that may be of significant use for tracking outcomes and ongoing clinical research to deliver better preventative interventions for PPH.

Postpartum haemorrhage trends in Sweden using the Robson ten group classification system: a population-based cohort study

OBJECTIVE

To examine postpartum haemorrhage (PPH) trends in Sweden using the Robson classification system.

DESIGN

Population-based cohort study.

SETTING

Sweden.

POPULATION

Deliveries in 2000-2016, classified as Robson groups 1-5 (singleton pregnancies in vertex presentation, from gestational weeks 37+0; n = 1 590 178).

METHODS

We examined temporal trends in PPH between 2000 and 2016 overall, and within each Robson group, and performed logistic regression to examine the influence of changes in risk factors (maternal, comorbidity, obstetric practice and infant factors) over time.

MAIN OUTCOME MEASURES

Postpartum haemorrhage, defined as an estimated blood loss of >1000 ml.

RESULTS

The overall PPH rate increased from 5.4 to 7.3%, corresponding to a 37% (OR 1.37, 95% CI 1.32-1.42) increase over time. Rates varied between Robson groups, ranging from 4.5% in group 3 to 14.3% in group 4b. Increasing trends in PPH were found in all Robson groups except for groups 2b and 4b (prelabour caesarean deliveries). In the unstratified analysis, adjusting for maternal, comorbidity and obstetric practice factors slightly attenuated the risk of PPH in the later period (2013-2016), compared with the reference period (2000-2004; crude OR 1.26, 95% CI 1.24-1.29, adjusted OR 1.22, 95% CI 1.20-1.25). Within individual Robson groups, changes in risk factors did not explain increasing rates of PPH.

CONCLUSIONS

Postpartum haemorrhage rates varied between Robson groups. Changes in risk factors could not explain the 37% increase in PPH for women in Robson groups 1-5 in Sweden, 2000-2016.

TWEETABLE ABSTRACT

Changes in risk factors could not explain the increasing trend of PPH in Sweden, and rates of PPH varied widely between Robson groups.

The association between postpartum hemorrhage and postpartum depression: A Swedish national register-based study

Background

Postpartum hemorrhage is an important cause of maternal death and morbidity. However, it is unclear whether women who experience postpartum hemorrhage are at an increased risk of postpartum depression.

Objectives

To examine whether postpartum hemorrhage is associated with postpartum depression.

Methods

We conducted a national register-based cohort study of 486,476 Swedish-born women who had a singleton livebirth between 2007 and 2014. We excluded women with pre-existing depression or who filled a prescription for an antidepressant before childbirth. We classified postpartum depression up to 12 months after giving birth by the presence of an International Classification of Diseases, version 10 (ICD-10) diagnosis code for depression or a filled outpatient prescription for an antidepressant. We used Cox proportional hazard models, adjusting for maternal sociodemographic and obstetric factors.

Results

Postpartum depression was identified in 2.0% (630/31,663) of women with postpartum hemorrhage and 1.9% (8601/455,059) of women without postpartum hemorrhage. In our unadjusted analysis, postpartum hemorrhage was not associated with postpartum depression (unadjusted hazard ratio (HR) = 1.06, 95% confidence interval (CI) 0.97–1.15). After adjusting for maternal age, parity, education, cohabitation status, maternal smoking status, and early pregnancy maternal BMI, gestational age, and birthweight, the association did not appreciably change, with confidence intervals overlapping the null (adjusted HR = 1.08, 95% CI 0.99, 1.17).

Conclusions

Within a population-based cohort of singleton women in Sweden with no prior history of depression, postpartum hemorrhage was not associated with postpartum depression.

Association of Gestational Age with Postpartum Hemorrhage: An International Cohort Study

BACKGROUND

Risk factors for postpartum hemorrhage, such as chorioamnionitis and multiple gestation, have been identified in previous epidemiologic studies. However, existing data describing the association between gestational age at delivery and postpartum hemorrhage are conflicting. The aim of this study was to assess the association between gestational age at delivery and postpartum hemorrhage.

METHODS

The authors conducted a population-based retrospective cohort study of women who underwent live birth delivery in Sweden between 2014 and 2017 and in California between 2011 and 2015. The primary exposure was gestational age at delivery. The primary outcome was postpartum hemorrhage, classified using International Classification of Diseases, Ninth Revision-Clinical Modification codes for California births and a blood loss greater than 1,000 ml for Swedish births. The authors accounted for demographic and obstetric factors as potential confounders in the analyses.

RESULTS

The incidences of postpartum hemorrhage in Sweden (23,323/328,729; 7.1%) and in California (66,583/2,079,637; 3.2%) were not comparable. In Sweden and California, the incidence of postpartum hemorrhage was highest for deliveries between 41 and 42 weeks' gestation (7,186/75,539 [9.5%] and 8,921/160,267 [5.6%], respectively). Compared to deliveries between 37 and 38 weeks, deliveries between 41 and 42 weeks had the highest adjusted odds of postpartum hemorrhage (1.62 [95% CI, 1.56 to 1.69] in Sweden and 2.04 [95% CI, 1.98 to 2.09] in California). In both cohorts, the authors observed a nonlinear (J-shaped) association between gestational age and postpartum hemorrhage risk, with 39 weeks as the nadir. In the sensitivity analyses, similar findings were observed among cesarean deliveries only, when postpartum hemorrhage was classified only by International Classification of Diseases, Tenth Revision-Clinical Modification codes, and after excluding women with abnormal placentation disorders.

CONCLUSIONS

The postpartum hemorrhage incidence in Sweden and California was not comparable. When assessing a woman's risk for postpartum hemorrhage, clinicians should be aware of the heightened odds in women who deliver between 41 and 42 weeks' gestation.

EDITOR’S PERSPECTIVE

Accuracy of identifying hospital acquired venous thromboembolism by administrative coding: implications for big data and machine learning research

Big data analytics research using heterogeneous electronic health record (EHR) data requires accurate identification of disease phenotype cases and controls. Overreliance on ground truth determination based on administrative data can lead to biased and inaccurate findings. Hospital-acquired venous thromboembolism (HA-VTE) is challenging to identify due to its temporal evolution and variable EHR documentation. To establish ground truth for machine learning modeling, we compared accuracy of HA-VTE diagnoses made by administrative coding to manual review of gold standard diagnostic test results. We performed retrospective analysis of EHR data on 3680 adult stepdown unit patients identifying HA-VTE. International Classification of Diseases, Ninth Revision (ICD-9-CM) codes for VTE were identified. 4544 radiology reports associated with VTE diagnostic tests were screened using terminology extraction and then manually reviewed by a clinical expert to confirm diagnosis. Of 415 cases with ICD-9-CM codes for VTE, 219 were identified with acute onset type codes. Test report review identified 158 new-onset HA-VTE cases. Only 40% of ICD-9-CM coded cases (n = 87) were confirmed by a positive diagnostic test report, leaving the majority of administratively coded cases unsubstantiated by confirmatory diagnostic test. Additionally, 45% of diagnostic test confirmed HA-VTE cases lacked corresponding ICD codes. ICD-9-CM coding missed diagnostic test-confirmed HA-VTE cases and inaccurately assigned cases without confirmed VTE, suggesting dependence on administrative coding leads to inaccurate HA-VTE phenotyping. Alternative methods to develop more sensitive and specific VTE phenotype solutions portable across EHR vendor data are needed to support case-finding in big-data analytics.

Hospital discharge data is not accurate enough to monitor the incidence of postpartum hemorrhage

INTRODUCTION

Postpartum hemorrhage remains a leading cause of maternal morbidity and mortality worldwide. Therefore, cumulative incidence of postpartum hemorrhage and severe postpartum hemorrhage are commonly monitored within and compared across maternity hospitals or countries for obstetrical safety improvement. These indicators are usually based on hospital discharge data though their accuracy is seldom assessed. We aimed to measure postpartum hemorrhage and severe postpartum hemorrhage using electronic health records and hospital discharge data separately and compare the detection accuracy of these methods to manual chart review, and to examine the temporal trends in cumulative incidence of these potentially avoidable adverse outcomes.

MATERIALS AND METHODS

We analyzed routinely collected data of 7904 singleton deliveries from a large Swiss university hospital for a three year period (2014-2016). We identified postpartum hemorrhage and severe postpartum hemorrhage in electronic health records by text mining discharge letters and operative reports and calculating drop in hemoglobin from laboratory tests. Diagnostic and procedure codes were used to identify cases in hospital discharge data. A sample of 334 charts was reviewed manually to provide a reference-standard and evaluate the accuracy of the other detection methods.

RESULTS

Sensitivities of detection algorithms based on electronic health records and hospital discharge data were 95.2% (95% CI: 92.6% 97.8%) and 38.2% (33.3% to 43.0%), respectively for postpartum hemorrhage, and 87.5% (85.2% to 89.8%) and 36.2% (26.3% to 46.1%) for severe postpartum hemorrhage. Postpartum hemorrhage cumulative incidence based on electronic health records decreased from 15.6% (13.1% to 18.2%) to 8.5% (6.7% to 10.5%) from the beginning of 2014 to the end of 2016, with an average of 12.5% (11.8% to 13.3%). The cumulative incidence of severe postpartum hemorrhage remained at approximately 4% (3.5% to 4.4%). Hospital discharge data-based algorithms provided significantly underestimated incidences.

CONCLUSIONS

Hospital discharge data is not accurate enough to assess the incidence of postpartum hemorrhage at hospital or national level. Instead, automated algorithms based on structured and textual data from electronic health records should be considered, as they provide accurate and timely estimates for monitoring and improvement in obstetrical safety. Furthermore, they have the potential to better code for postpartum hemorrhage thus improving hospital reimbursement.

Severity of shock, rate of physiological stabilization and organ failure in healthy women admitted to the intensive care unit following major peripartum hemorrhage: A retrospective, descriptive study

PURPOSE

To describe shock severity, physiological stabilization and organ failure in healthy women admitted to the intensive care unit (ICU) after major peripartum hemorrhage (PPH).

MATERIALS AND METHODS

Retrospective, descriptive, single center study.

RESULTS

Twenty-nine women median age 33 years (interquartile range [IQR] 30-36) and gravidity 5 pregnancies (IQR 3-9) were studied. One woman died. The median maternal admission hematocrit was 28.8 (IQR 25.7-32.4). Median transfusion rates were nine units of packed red blood cells (IQR 7-12.25), eight fresh frozen plasma (IQR 6-12), 17 platelets (IQR 10-22) and 15 cryoprecipitate (IQR 9.75-20). Blood pressure dropped significantly in the six hours following ICU admission. Nonetheless, lactate decreased from 3.23 mmol/L to 1.54 mmol/L within 24 h of ICU admission, renal and pulmonary function were unaffected and coagulopathy was never observed. Two-thirds of the women underwent hysterectomy. One-third underwent repeated surgery. The median length of ICU stay was <48 h and that of mechanical ventilation was <24 h. Increased transfusion rates correlated with lengthier ICU admission (p ≤ 0.01 regardless of blood product).

CONCLUSIONS

Ongoing hemorrhage in women with severe PPH manifests subtly and often requires active intervention. Hemorrhage control is required to achieve physiological stabilization and minimize organ damage.

Association between ionised calcium and severity of postpartum haemorrhage: a retrospective cohort study

BACKGROUND

Postpartum haemorrhage (PPH) is often complicated by impaired coagulation. We aimed to determine whether the level of ionised calcium (Ca²⁺), an essential coagulation co-factor, at diagnosis of PPH is associated with bleeding severity.

METHODS

This was a retrospective cohort study of women diagnosed with PPH during vaginal delivery between January 2009 and April 2020. Ca²⁺ levels at PPH diagnosis were compared between women who progressed to severe PPH (primary outcome) and those with less severe bleeding. Severe PPH was defined by transfusion of ≥2 blood units, arterial embolisation or emergency surgery, admission to ICU, or death. Associations between other variables (e.g. fibrinogen concentration) and bleeding severity were also assessed.

RESULTS

For 436 patients included in the analysis, hypocalcaemia was more common among patients with severe PPH (51.5% vs 10.6%, P<0.001). In a multivariable logistic regression model, Ca²⁺ and fibrinogen were the only parameters independently associated with PPH severity with odds ratios of 1.14 for each 10 mg dl^-1 decrease in fibrinogen (95% confidence interval [CI], 1.05-1.24; P=0.002) and 1.97 for each 0.1 mmol L^-1 decrease in Ca²⁺ (95% CI, 1.25-3.1; P=0.003). The performance of Ca²⁺ or fibrinogen was not significantly different (area under the curve [AUC]=0.79 [95% CI, 0.75-0.83] vs AUC=0.86 [95% CI, 0.82-0.9]; P=0.09). The addition of Ca²⁺ to fibrinogen improved the model, leading to AUC of 0.9 (95% CI, 0.86-0.93), P=0.03.

CONCLUSIONS

Ca²⁺ level at the time of diagnosis of PPH was associated with risk of severe bleeding. Ca²⁺ monitoring may facilitate identification and treatment of high-risk patients.

Accuracy of postpartum hemorrhage coding in the Swedish Pregnancy Register

INTRODUCTION

Postpartum hemorrhage (PPH) is recognized as a leading cause of obstetric morbidity and mortality. Population-wide studies have used International Classification of Diseases (ICD) diagnostic codes to track and report the prevalence of PPH. Although the 10th revision (ICD-10) was introduced in Sweden in 1997, the accuracy of ICD-10 codes for PPH is not known. Thus, the aim was to determine the accuracy of diagnostic coding for PPH in the Swedish Pregnancy Register.

MATERIAL AND METHODS

We performed a retrospective cohort study of 609 807 deliveries in Sweden between 2014 and 2019. Information on ICD-10 codes for PPH and estimated blood loss were extracted from the Swedish Pregnancy Register. Using an estimated blood loss >1000 mL as the reference standard, we evaluated the diagnostic accuracy of ICD-10 codes for PPH by estimating sensitivity, specificity, positive predictive value and negative predictive value with exact binomial 95% confidence intervals (CIs). In our secondary analysis, we assessed the ICD-10 coding accuracy for severe PPH, defined as an estimated blood loss >1000 mL and transfusion of at least 1 unit of red blood cells registered in the Scandinavian Donations and Transfusion database.

RESULTS

Of the 609 807 deliveries, 43 312 (7.1%) had an ICD-10 code for PPH and 45 071 (7.4%) had an estimated blood loss >1000 mL. The ICD codes had a sensitivity of 88.5% (95% CI 88.2-88.7), specificity of 99.4% (95% CI 99.4-99.4), positive predictive value of 92.0% (95% CI 91.8-92.3) and negative predictive value of 99.1% (95% CI 99.1-99.1). In our secondary analysis, on deliveries with severe PPH, the sensitivity for an ICD code was 91.3% (95% CI 90.7-91.9), whereas specificity was 83.5% (95% CI 82.3-84.6).

CONCLUSIONS

Our findings indicate that ICD-10 codes for PPH in Sweden have moderately high sensitivity and excellent specificity. These results suggest that PPH diagnostic codes in medical records and linked pregnancy and birth registers can be used for research, quality improvement and reporting PPH prevalence in Sweden.

Obstetric haemorrhage and risk of cardiovascular disease after three decades: a population-based cohort study

OBJECTIVE

To investigate the association between obstetric haemorrhage and cardiovascular disease up to three decades after pregnancy.

DESIGN

Population-based cohort study.

SETTING AND POPULATION

All women who delivered between 1989 and 2016 in Quebec, Canada.

METHODS

Using hospital admissions data, 1 224 975 women were followed from their first delivery until March 2018. The main exposure measures were antenatal (placenta praevia, placental abruption, peripartum haemorrhage) or postpartum haemorrhage, with or without transfusion. Adjusted Cox regression models were used to assess the association between obstetric haemorrhage and future cardiovascular disease.

MAIN OUTCOME MEASURE

Cardiovascular hospitalisation.

RESULTS

Among 104 291 (8.5%) women with haemorrhage, 4612 (4.4%) required transfusion. Women with haemorrhage had a higher incidence of cardiovascular hospitalisation than women without haemorrhage (15.5 versus 14.1 per 10 000 person-years; 2437 versus 28 432 events). Risk of cardiovascular hospitalisation was higher for obstetric haemorrhage, with or without transfusion, compared with no haemorrhage (adjusted hazard ratio [aHR] 1.06, 95% CI 1.02-1.10). Women with haemorrhage and transfusion had a substantially greater risk of cardiovascular hospitalisation (aHR 1.47, 95% CI 1.23-1.76). Among transfused women, placental abruption (aHR 1.79, 95% CI 1.06-3.00) and postpartum haemorrhage (aHR 1.38, 95% CI 1.13-1.68) were both associated with risk of cardiovascular hospitalisation. Antenatal haemorrhage with transfusion was associated with 2.46 times the risk of cardiovascular hospitalisation at 5 years (95% CI 1.59-3.80) and 2.14 times the risk at 10 years (95% CI 1.47-3.12).

CONCLUSIONS

Obstetric haemorrhage requiring transfusion is associated with maternal cardiovascular disease. The benefit of cardiovascular risk prevention in pregnant women with obstetric haemorrhage requires further investigation.

TWEETABLE ABSTRACT

Risk of future cardiovascular disease is increased for women with obstetric haemorrhage who require transfusion.

A principled framework for phenotyping postpartum hemorrhage across multiple levels of severity

Maternal morbidity and mortality have gained major attention recently, spurred on by rising domestic rates even as maternal mortality decreases in Europe. A major driver of morbidity and mortality among delivering women is postpartum hemorrhage (PPH). PPH is currently phenotyped using the subjective measure of 'Estimated blood loss' (EBL), which has been shown to be unreliable for tracking quality. Here we present a framework for phenotyping PPH into multiple severity levels, using a combination of data-driven techniques and expert-derived clinical indicators. We validate the framework by predicting large drops in hematocrit and quantitative blood loss, finding that the framework performs better in predicting coded PPH than a hematocrit-based predictor or predictors based on other metrics such as blood transfusions, and does better in predicting quantitative blood loss, a gold standard metric for blood loss that we have for a subset of patients, than any predictor we could build using hematocrit drops alone. In all, we present a principled framework that can be used to phenotype PPH in hospitals using readily available EHR data, and that will perform with more granularity and accuracy than existing methods.