Main Article Content
Hospital datasets are a valuable resource for examining prevalence and outcomes of medical conditions during pregnancy. To enable effective research and health planning, it is important to determine whether variables are reliably captured.
To examine the reliability of reporting of gestational and pre-existing diabetes, hypertension, thyroid conditions, and morbid obesity in coded hospital records that inform the population-level New South Wales Admitted Patient Data Collection.
Coded hospital admission data from two large tertiary hospitals in New South Wales, from 2011 to 2015, were compared with obstetric data, collected by midwives at outpatient pregnancy booking and in hospital after birth, as the reference standard. Records were deterministically linked and sensitivity, specificity, positive predictive values and negative predictive values for the conditions of interest were obtained.
There were 36,051 births included in the analysis. Sensitivity was high for gestational diabetes (83.6%, 95% CI 82.4–84.7%), pre-existing diabetes (88.2%, 95% CI 84.1–91.6%), and gestational hypertension (80.1%, 95% CI 78.2–81.9%), moderate for chronic hypertension (53.5%, 95% CI 47.8–59.1%), and low for thyroid conditions (12.9%, 95% CI 11.7–14.2%) and morbid obesity (9.8%, 95% CI 7.6–12.4%). Specificity was high for all conditions (≥97.8%, 95% CI 97.7–98.0) and positive predictive value ranged from 53.2% for chronic hypertension (95% CI 47.5–58.8%) to 92.7% for gestational diabetes (95% CI 91.8–93.5%).
Our findings suggest that coded hospital data are a reliable source of information for gestational and pre-existing diabetes and gestational hypertension. Chronic hypertension is less consistently reported, which may be remedied by grouping hypertension types. Data on thyroid conditions and morbid obesity should be used with caution, and if possible, other sources of data for those conditions should be sought.
Timely, high quality, clinically relevant obstetric research is important to improve patient care for women and their newborns, so that as often as possible new mothers leave hospital healthy and with a healthy newborn. Hospital admission data are an efficient and rich resource for conducting such research on conditions that affect pregnancy and its outcomes [1–4]. Hospital data are also vital to inform decisions about health planning and, in many places, hospital funding . Obstetric data do not always capture the range of procedures and conditions that can affect pregnant women and their babies, or may capture them in a way that is more difficult to analyse, such as free text. In addition, population level obstetric data are less commonly available than hospital data. Hospital data are usually coded with diagnosis and procedure codes following international coding standards, and therefore provide a useful alternative or supplement to obstetric data for obstetric research [1–4, 6]. In order to perform effective population-based research, however, researchers need information on the extent to which the data accurately reflect the clinical situation.
Reporting accuracy of diagnoses and procedures in hospital data may be affected by changes in practice and as different conditions become the focus of guidelines, management and audits. In the case of gestational diabetes, for example, changes in thresholds for diagnosis following the publication of results of a large prospective blinded observational study  have resulted in rates more than doubling in New South Wales (NSW), from 5.7% in 2005 to 12.2% in 2014, and gestational diabetes now accounts for almost 30% of planned births before 39 weeks gestation (unpublished data). Hospital data, collected when a woman is admitted to a maternity facility (hospital or birth centre) during pregnancy and birth, are a valuable resource for examining such changes over time and their impacts on outcomes, however they are typically collected for administrative purposes such as billing, rather than for research. Validation is therefore important to assess the reliability of hospital data sources.
This study examines reporting of maternal medical conditions including diabetes, hypertension, thyroid conditions and morbid obesity (BMI>40kg/m2), which are associated with increased risks of adverse outcomes for mothers and babies [8–11], in hospital data. The most recent validation study assessing reporting of diabetes in coded NSW hospital data showed moderate reliability for gestational and high reliability for pre-existing diabetes, however it assessed data from 2002 , and predates the changes in diagnostic criteria. Previous studies have shown variable sensitivity for chronic hypertension (ranging from 44%–86%) [13, 14], gestational hypertension (10–71%) , thyroid conditions (10–97%)  and a systematic review  found only a single validation study for morbid obesity in hospital data, which showed low sensitivity (10%) . Further, clinical practice and the obstetric population may have changed in the intervening years.
Validation studies may be conducted comparing coded hospital data with medical charts, however this process is time consuming and expensive to conduct, and tends to reflect only a short time period. An alternative approach is to compare reporting in two independent databases [17–19]. Here we compare reporting in coded hospital data, taken from the hospital records of birth and any pregnancy admissions, to obstetric data, collected from antenatal clinics and the obstetric record of the pregnancy and birth, using obstetric data as the reference standard.
The coded hospital data are drawn from the electronic medical record for a hospital admission, extracted by trained clinical coders following the International Classification of Diseases (ICD). A primary purpose of the coded diagnoses is to enable activity based funding (whereby hospitals are provided government funding largely on the basis of the treatment of patients), healthcare management and planning. The Australian Coding Standards provides guidelines to ensure consistency in clinical coding nationally. These limit conditions coded to those affecting patient management in the current admission, meaning that some chronic or other present conditions that do not affect the admission are not recoded. The guidelines also require coding to be substantiated by clear medical record documentation or confirmation from a clinician, and prohibit interpretation of results; for example, a recorded BMI of 40.1 kg/m2 may not be assigned a code for obesity without an explicit, documented diagnosis of obesity in the notes [20, 21]. Coded data also inform various healthcare management and planning purposes, including the population-level New South Wales Admitted Patient Data Collection.
Obstetrics data were drawn from ObstetriX, a clinical database administered during pregnancy, birth and the early postnatal period, collected by midwives and partially self-reported. ObstetriX data inform the population-level New South Wales Perinatal Data Collection. ObstetriX may be considered an imperfect reference standard due to the different purposes and perspectives of the databases, however most reference standards are not without error and uncertainty [22, 23], and large population-level datasets have been demonstrated to be robust to random errors and omissions .
Women who gave birth to singleton infants in two tertiary hospitals in the Sydney metropolitan area between 1 January 2011 and 31 December 2015 formed the study population. All births in a hospital or birth centre, which in NSW are publicly funded facilities associated with public hospitals, are considered inpatient admissions and are assigned both an electronic medical record and obstetric record (in ObstetriX or a similar database such as eMaternity). In NSW, 99% of women birth in a hospital or birth centre . Women who had prearranged to give birth in a different hospital to the actual hospital of birth were excluded, because antenatal data would have been collected at their hospital of booking rather than the hospital of birth, and therefore ObstetriX records may be incomplete for these women. Births of ≥28 weeks gestation were included for diabetes, as the screening test for gestational diabetes is recommended to be completed by 28 weeks, and ≥24 weeks gestation for other conditions.
Obstetric data were obtained from the ObstetriX system (Meridian Health Informatics, Sydney, Australia), and linked to hospital data for admissions during pregnancy and the birth admission, drawn from the electronic medical record.
ObstetriX contains maternal health and demographic data, obstetric history and pregnancy details primarily obtained initially at the face-to-face booking consultation with a midwife (an outpatient encounter, by 16 weeks gestation). It is updated with labour, birth and postnatal information obtained during the birth admission, which is recorded and entered into the system contemporaneously. Data are recorded in checkboxes or drop-down menus, with a small number of free text fields. The data are entered by midwives, who do not have access to the hospital codes, as coding is performed four to six weeks after discharge. Most procedures and conditions are recorded as present, absent or unknown/missing. Presence or absence of the conditions of interest were obtained from checkbox or dropdown variables for the conditions, with the exception of morbid obesity, which was defined as having a body mass index (BMI) of 40.0 kg/m2 or above in ObstetriX. A subset of ObstetriX data is submitted electronically to form the state-wide New South Wales Perinatal Data Collection.
The electronic medical record contains a record of diagnoses coded according to the International Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification (ICD-10-AM), with a small number coded using Systematized Nomenclature of Medicine – Clinical Terms (SNOMED CT) codes. Procedures are coded following the Australian Classification of Health Interventions, Eighth Edition (ACHI). Coding is performed by medical coders based on clinical documentation in the electronic medical record. Although coders have access to ObstetriX data collected at birth, this information is not always coded unless the diagnoses and procedures are clearly present in the electronic medical record. Medical record diagnosis and procedure codes for admissions throughout pregnancy and the birth admission were searched for the relevant conditions. Where no record of the condition was found in any admission, the condition was deemed to be ‘absent’ for the purpose of this study. The diagnosis and procedure codes used in this study are provided in Supplementary Table 1. In order to maximise sensitivity for morbid obesity, parent codes for obesity and overweight (E66) and localised adiposity (E65) were used. The coded hospital data are submitted electronically from each hospital to form the New South Wales Admitted Patient Data Collection.
Records from the two sources were deterministically linked using patient Medical Record Number and checked using other personal identifiers, by personnel external to the project. Data were de-identified for analysis.
Reporting of gestational diabetes and other chronic disorders in the hospital data was compared to that in ObstetriX, using ObstetriX as the reference standard. Conditions recorded in any pregnancy admission or the birth admission in the coded hospital data were compared to conditions recorded in ObstetriX at any time for that pregnancy and birth. Records with missing data were excluded from the analysis of that variable. Sensitivity, specificity, positive and negative predictive values are reported with exact confidence intervals. Sensitivity and specificity were also calculated separately for the two hospitals. Analyses were performed in SAS 9.3. This study received ethics approval from the Northern Sydney Local Health District Human Research Ethics Committee (LNR/17/HAWKE32).
There were 38,343 singleton infants born at 24 weeks gestation and older at the two hospitals between January 2011 and December 2015 (Figure 1). Of these, 36,051 received antenatal care at their birth hospital and were included in the analysis. There were 35,928 infants born at 28 weeks and older and included in the analysis of gestational and pre-existing diabetes.
Sensitivity, specificity, positive and negative predictive values for the conditions of interest are provided in Table 1. Gestational diabetes was reliably reported over the period (sensitivity 83.6%, specificity 99.2%, PPV 92.7%), as was pre-existing diabetes (sensitivity 88.2%, specificity 99.9%, PPV 86.0%). Pre-eclampsia and eclampsia, gestational hypertension and any hypertension had good sensitivity (80.0%, 80.1%, 81.5%, respectively), but positive predictive value was moderate, showing that 30-40% of cases reported in the hospital data were not reported in ObstetriX (PPV 59.7%, 65.6%, 70.4%, respectively), and sensitivity and PPV for chronic hypertension were lower (sensitivity 53.5%, PPV 53.2%). Reporting of morbid obesity and thyroid conditions showed very low sensitivity (9.8%, 12.9% respectively), with moderate to high PPV (65.6% and 82.3%, respectively). Sensitivity increased for gestational diabetes, from 60.5% in 2011 to 95.7% in 2015 (Figure 2, Table 2), and increased less dramatically for pre-existing and any diabetes, gestational hypertension and any hypertension. Specificity and NPV were very high for all conditions examined.
|Condition||ObstetriX||Coded hospital data||Sensitivity % (95% CI)||Specificity % (95% CI)||PPV % (95% CI)||NPV % (95% CI)|
|Gestational diabetes||4055||3654||83.6 (82.4–84.7)||99.2 (99.0–99.3)||92.7 (91.8–93.5)||97.9 (97.8–98.1)|
|Pre-existing diabetes||306||314||88.2 (84.1–91.6)||99.9 (99.8–99.9)||86.0 (81.6–89.6)||99.9 (99.9–99.9)|
|Any diabetes||4361||3953||84.8 (83.7–85.9)||99.2 (99.1–99.3)||93.6 (92.8–94.3)||97.9 (97.8–98.1)|
|Morbid obesity (BMI>40kg/m2)||622||93||9.8 (7.6–12.4)||99.9 (99.9–99.9)||65.6 (55.0–75.1)||98.4 (98.3–98.5)|
|Thyroid conditions||2895||453||12.9 (11.7–14.2)||99.8 (99.7–99.8)||82.3 (78.5–85.7)||92.9 (92.6–93.2)|
|Chronic hypertension||314||316||53.5 (47.8–59.1)||99.6 (99.5–99.6)||53.2 (47.5–58.8)||99.6 (99.5–99.7)|
|Pre-eclampsia and eclampsia||610||817||80.0 (76.6–83.1)||99.1 (99.0–99.2)||59.7 (56.3–63.1)||99.7 (99.6–99.7)|
|Gestational hypertension1||1778||2170||80.1 (78.2–81.9)||97.8 (97.7–98.0)||65.6 (63.6–67.6)||99.0 (98.8–99.1)|
|Any hypertension||2035||2357||81.5 (79.8–83.2)||97.9 (97.8–98.1)||70.4 (68.5–72.2)||98.9 (98.8–99.0)|
|Condition||Characteristic||Value||ObstetriX||Hospital data||Sensitivity % (95% CI)||Specificity % (95% CI)||PPV % (95% CI)||NPV % (95% CI)|
|Gestational diabetes||Parity||Nulliparous||1893||1749||86.9 (85.3–88.4)||99.3 (99.1–99.4)||94.1 (92.8–95.1)||98.3 (98.1–98.5)|
|Parous||2147||1889||80.5 (78.7–82.1)||99.0 (98.9–99.2)||91.5 (90.1–92.7)||97.5 (97.3–97.8)|
|Year of birth||2011||653||422||60.5 (56.6–64.3)||99.6 (99.4–99.7)||93.6 (90.8–95.7)||96.0 (95.5–96.5)|
|2012||690||522||69.4 (65.8–72.8)||99.3 (99.1–99.5)||91.8 (89.1–94.0)||96.8 (96.3–97.2)|
|2013||764||714||85.9 (83.2–88.3)||99.1 (98.8–99.3)||91.9 (89.6–93.8)||98.3 (97.9–98.6)|
|2014||918||941||95.0 (93.4–96.3)||99.0 (98.7–99.2)||92.7 (90.8–94.3)||99.3 (99.1–99.5)|
|2015||1030||1055||95.7 (94.3–96.9)||98.9 (98.6–99.1)||93.5 (91.8–94.9)||99.3 (99.1–99.5)|
|Model of care||Hospital medical||2709||2454||85.0 (83.7–86.4)||98.6 (98.4–98.9)||93.9 (92.9–94.8)||96.4 (96.1–96.8)|
|Midwife||1020||914||80.7 (78.1–83.1)||99.5 (99.4–99.6)||90.0 (87.9–91.9)||98.9 (98.7–99.0)|
|Private OB or GP||326||285||80.1 (75.3–84.3)||99.1 (98.7–99.5)||91.6 (87.7–94.5)||97.7 (97.1–98.2)|
|Pre-existing diabetes||Parity||Nulliparous||117||126||91.5 (84.8–95.8)||99.9 (99.8–99.9)||84.9 (77.5–90.7)||99.9 (99.9, 100.0)|
|Parous||188||187||86.2 (80.4–90.8)||99.9 (99.8–99.9)||86.6 (80.9–91.2)||99.9 (99.8–99.9)|
|Year of birth||2011||35||34||77.1 (59.9–89.6)||99.9 (99.8–100.0)||79.4 (62.1–91.3)||99.9 (99.8–99.9)|
|2012||58||56||81.0 (68.6–90.1)||99.9 (99.8–99.9)||83.9 (71.7–92.4)||99.8 (99.7–99.9)|
|2013||62||63||83.9 (72.3–92.0)||99.8 (99.7–99.9)||82.5 (70.9–90.9)||99.9 (99.7–99.9)|
|2014||72||76||94.4 (86.4–98.5)||99.9 (99.8–100.0)||89.5 (80.3–95.3)||99.9 (99.9–100.0)|
|2015||79||85||96.2 (89.3–99.2)||99.9 (99.8–99.9)||89.4 (80.8–95.0)||100.0 (99.9–100.0)|
|Model of care||Hospital medical||261||266||89.7 (85.3–93.1)||99.8 (99.7–99.8)||88.0 (83.4–91.6)||99.8 (99.7–99.9)|
|Midwife||11||13||54.5 (23.4–83.3)||100.0 (99.9–100.0)||46.2 (19.2–74.9)||100.0 (99.9–100.0)|
|Private OB or GP||34||35||88.2 (72.5–96.7)||99.8 (99.6–99.9)||85.7 (69.7–95.2)||99.9 (99.7–100.0)|
|Management2||Type 1 insulin||141||130||92.2 (86.5–96.0)||–||–||–|
|Type 2 insulin||94||85||90.4 (82.6–95.5)||–||–||–|
|Type 2 no insulin||71||55||77.5 (66.0–86.5)||–||–||–|
|Any diabetes||Parity||Nulliparous||2010||1872||88.0 (86.5–89.4)||99.3 (99.1–99.4)||94.4 (93.3–95.4)||98.4 (98.2–98.6)|
|Parous||2335||2064||82.0 (80.4–83.6)||99.1 (98.9–99.2)||92.8 (91.6–93.9)||97.5 (97.3–97.7)|
|Year of birth||2011||688||458||62.5 (58.8–66.1)||99.5 (99.3–99.7)||93.9 (91.3–95.9)||96.0 (95.4–96.4)|
|2012||748||576||71.8 (68.4–75.0)||99.4 (99.2–99.6)||93.2 (90.9–95.1)||96.7 (96.3–97.2)|
|2013||826||775||87.2 (84.7–89.4)||99.1 (98.8–99.3)||92.9 (90.9–94.6)||98.3 (97.9–98.6)|
|2014||990||1009||95.5 (94.0–96.7)||99.0 (98.7–99.2)||93.7 (92.0–95.1)||99.3 (99.1–99.5)|
|2015||1109||1135||96.2 (94.9–97.3)||98.9 (98.6–99.2)||94.0 (92.5–95.3)||99.3 (99.1–99.5)|
|Model of care||Hospital medical||2970||2705||86.5 (85.3–87.7)||98.8 (98.5–99.0)||95.0 (94.1–95.8)||96.4 (96.0–96.7)|
|Midwife||1031||926||81.0 (78.5–83.3)||99.5 (99.4–99.6)||90.2 (88.1–92.0)||98.9 (98.7–99.0)|
|Private OB or GP||360||321||81.7 (77.3–85.5)||99.0 (98.6–99.4)||91.6 (88.0–94.4)||97.7 (97.0–98.2)|
|Any hypertension||Parity||Nulliparous||1139||1372||86.6 (84.4–88.5)||97.5 (97.3–97.8)||71.9 (69.4–74.2)||99.0 (98.8–99.2)|
|Parous||889||978||75.1 (72.2–78.0)||98.3 (98.1–98.5)||68.3 (65.3–71.2)||98.8 (98.6–98.9)|
|Year of birth||2011||420||414||75.2 (70.8–79.3)||98.5 (98.2–98.8)||76.3 (71.9–80.3)||98.4 (98.1–98.7)|
|2012||395||448||79.5 (75.2–83.4)||98.0 (97.6–98.3)||70.1 (65.6–74.3)||98.8 (98.5–99.0)|
|2013||384||440||81.8 (77.5–85.5)||98.1 (97.8–98.4)||71.4 (66.9–75.5)||98.9 (98.7–99.2)|
|2014||427||536||83.6 (79.7–87.0)||97.5 (97.1–97.8)||66.6 (62.4–70.6)||99.0 (98.7–99.2)|
|2015||409||519||87.5 (83.9–90.6)||97.7 (97.3–98.0)||69.0 (64.8–72.9)||99.3 (99.0–99.4)|
|Model of care||Hospital medical||1310||1424||82.9 (80.8–84.9)||97.3 (97.0–97.6)||76.3 (74.0–78.5)||98.2 (98.0–98.4)|
|Midwife||468||681||81.0 (77.1–84.4)||98.4 (98.2–98.5)||55.7 (51.8–59.4)||99.5 (99.4–99.6)|
|OB or GP||232||246||81.5 (75.9–86.2)||98.1 (97.5–98.5)||76.8 (71.0–82.0)||98.5 (98.0–98.9)|
|Gestational hypertension1||Parity||Nulliparous||1041||1284||85.5 (83.2–87.6)||97.5 (97.3–97.7)||69.3 (66.7–71.8)||99.0 (98.9–99.2)|
|Parous||731||879||72.4 (69.0–75.6)||98.1 (97.9–98.3)||60.2 (56.9–63.4)||98.9 (98.7–99.0)|
|Year of birth||2011||364||384||75.3 (70.5–79.6)||98.3 (98.0–98.6)||71.4 (66.5–75.8)||98.6 (98.3–98.9)|
|2012||356||409||77.5 (72.8–81.8)||98.0 (97.7–98.3)||67.5 (62.7–72.0)||98.8 (98.5–99.1)|
|2013||324||404||80.6 (75.8–84.7)||97.9 (97.5–98.2)||64.6 (59.7–69.3)||99.1 (98.8–99.3)|
|2014||371||492||82.5 (78.2–86.2)||97.4 (97.0–97.8)||62.2 (57.7–66.5)||99.1 (98.8–99.3)|
|2015||363||481||84.6 (80.4–88.1)||97.5 (97.1–97.9)||63.8 (59.4–68.1)||99.2 (99.0–99.4)|
|Model of care||Hospital medical||1106||1277||80.8 (78.4–83.1)||97.0 (96.7–97.3)||70.0 (67.4–72.5)||98.3 (98.1–98.5)|
|Midwife||451||662||81.4 (77.5–84.9)||98.4 (98.2–98.6)||55.4 (51.6–59.3)||99.5 (99.4–99.6)|
|Private OB or GP||196||225||80.6 (74.4–85.9)||97.8 (97.2–98.3)||70.2 (63.8–76.1)||98.7 (98.2–99.1)|
Sensitivity and specificity were similar between the two hospitals (Supplementary Table 2). Specificity was very high for both hospitals. Sensitivity was slightly higher at Hospital One for non-diabetes related conditions, and there was less variation in sensitivity between hospitals for any hypertension compared to specific types of hypertension.
Consistency of reporting by characteristics of the pregnancy is provided in Table 2. For all conditions examined in detail, sensitivity was higher when a woman was nulliparous and where there was a hospital medical model of care (exclusive and with shared GP care), with the exception of gestational hypertension where midwife care had slightly higher sensitivity. Sensitivity to gestational diabetes was higher where the condition was managed with insulin or oral therapy compared to diet. Sensitivity increased by year for all conditions, most dramatically for gestational diabetes. Positive predictive value increased with time for pre-existing diabetes, but remained fairly stable for most other conditions, and decreased for hypertension. The median BMI of all those with a diagnosis in the hospital data of overweight, obesity or localised adiposity was 41.9 kg/m2 (IQR 37.9–46.9 kg/m2, n = 93).
The hospital data appeared to contain some misclassification of pre-existing conditions as conditions that arose in pregnancy (Table 2). Among cases of gestational diabetes that were recorded in ObstetriX but uncoded in the hospital data (667 of 4055, 16.4%), 12 (1.8%) were recorded as having pre-existing diabetes, while among cases of pre-existing diabetes that were recorded in ObstetriX but uncoded in the hospital data (36 of 306, 5.2%), 19 (52.8%) were recorded as having gestational diabetes (6% of total cases). Similarly, of the gestational hypertension cases uncoded in the hospital data (354 of 1778, 19.9%), 44 (12.4%) were reported with chronic hypertension, while 79 (54.1%) of chronic hypertension cases uncoded in the hospital data (146 of 314, 46.5%, or 25% of total cases identified in ObstetriX) were reported instead to have gestational hypertension.
Rates of the conditions investigated at the two tertiary hospitals were generally within what would be expected for the population (Table 3). However morbid obesity was lower than expected, while gestational hypertension and thyroid conditions were higher than expected for the population.
|Variable||ObstetriXn (%)||Hospitaldata n (%)||Eithern (%)||Population prevalencein other studies %|
|Gestational diabetes||4055 (11.4)||3654 (10.2)||4335 (12.1)||6.5–13.82|
|Pre-existing diabetes||306 (0.9)||314 (0.9)||351 (1.0)||1.0–1.03|
|Any diabetes||4361 (12.2)||3953 (11.0)||4629 (12.9)||7.0–9.84|
|Morbid obesity (BMI >40 kg/m2)||622 (1.8)||93 (0.3)||654 (1.8)||3.0–3.05|
|Thyroid conditions||2895 (8.0)||453 (1.3)||2975 (8.3)||2.0–3.06|
|Chronic hypertension||314 (0.9)||316 (0.9)||462 (1.3)||0.8–0.84|
|Pre-eclampsia and eclampsia||610 (1.7)||817 (2.3)||939 (2.6)||1.5–1.74|
|Gestational hypertension1||1778 (4.9)||2170 (6.0)||2524 (7.0)||2.8–3.14|
|Any hypertension||2035 (5.6)||2357 (6.5)||2733 (7.6)||5.1–5.64|
This study examined the reliability of coded hospital data for reporting of gestational diabetes and other maternal conditions, compared with ObstetriX data as the reference standard. Diabetes was well reported in the hospital data overall, with an increase over the time period from moderate to very high accuracy. Gestational hypertension, pre-eclampsia and eclampsia were moderately well reported, with high sensitivity but moderate PPV, while chronic hypertension had only moderate sensitivity and PPV. Using a broad ‘any hypertension’ category increased sensitivity and PPV and should be considered for studies using these data. Thyroid conditions and morbid obesity were poorly reported, with very low sensitivity and morbid obesity also showing moderate ‘false’ positives among cases that were reported, suggesting caution in the use of these data. Other sources of data on those two conditions should be sought where possible.
Sensitivity observed for pre-existing diabetes and gestational diabetes at the start of the study period were similar to that reported elsewhere [12, 21]. Sensitivity for chronic hypertension was within the range [13, 14] or slightly higher than reported elsewhere , while for gestational hypertension it was higher than previously observed [15, 21]. High sensitivity and low PPV for pre-eclampsia was also consistent with previous work [13, 26]. For thyroid conditions, sensitivity was at the lower limit of the range previously reported  and for morbid obesity sensitivity was consistent with poor ascertainment reported elsewhere . The high specificity observed for all conditions was consistent with previous studies [12–16]. The higher sensitivity for more severe diabetes (among women with Type 1 or Type 2 requiring insulin compared to those not requiring insulin for pre-existing diabetes, and receiving insulin or oral therapy compared to those who did not for gestational diabetes), is consistent with previous research on diabetes  and findings that greater severity of a condition is associated with better reporting [17, 27].
Sensitivity increased over time for all conditions, although this was particularly stark for gestational diabetes and chronic hypertension. The general trend of increasing sensitivity over time may be partly related to the introduction of activity based funding, a government scheme for calculating hospital funding, which was ratified in 2011 and introduced in NSW in 2012. A systematic review of accuracy in UK hospital data found that a similar funding method, known as ‘results based funding’, was also associated with an increase in coding accuracy . Sensitivity for gestational diabetes increased dramatically from 61% in 2011 to 96% in 2015, which is likely related to the change in diagnostic criteria and corresponding increased incidence and clinical focus on the condition and its management. This may have had flow on effects for pre-existing diabetes, for which the trend mirrored gestational diabetes.
Inconsistent reporting of thyroid conditions and morbid obesity are a reflection of the difficulties in capturing pre-existing, non-acute conditions from coded hospital data. Government policy and coding standards generally allow only conditions that affect the current admission to be coded, with the exception of diabetes and some other specific conditions. If a thyroid condition is well controlled, it requires negligible hospital resources during admission and is therefore unlikely to be recorded, although it may be recorded in the neonatal record since it can be more likely to impact the infant than mother. Morbid obesity may not be perceived as relevant to most hospital admissions during pregnancy, or staff may be reluctant to label a woman as obese due to negative connotations. The different timing of data collection between the two datasets may also play a role; for example, a person may report a history of thyroid disorder that is recorded in ObstetriX but has resolved or no longer requires medication, and therefore is uncoded in the hospital data.
We found evidence of misclassification of pre-existing conditions as conditions arising in pregnancy, with a tendency for chronic hypertension and pre-existing diabetes to be reported as gestational hypertension and diabetes, which has been previously observed . This was more common for diabetes (25% of all cases) compared to hypertension (6% of all cases).
Rates reported in ObstetriX were similar to what would be expected for the population [29–34]. Slightly higher rates of gestational hypertension and thyroid conditions may reflect that the hospitals included are tertiary hospitals. Morbid obesity was lower than the expectation for the population, perhaps reflecting the fact that both hospitals are located in metropolitan Sydney, where morbid obesity rates tend to be slightly lower than rural and regional areas .
Consistency of reporting was similar for the two hospitals, suggesting that the results hold across different socioeconomic and ethnic patient populations, locations, and facilities. Slight differences in the rates may reflect the different demographic compositions, with Hospital One tending to have an older obstetric population with a different mix of ethnicities and comorbidities .
ObstetriX is an imperfect reference standard, given the different purposes, perspectives and collection times of the two datasets. ObstetriX data are largely self-reported, which may be inaccurate or inconsistent , and the data are entered by busy clinical staff, with accuracy of data entry sometimes difficult to achieve when personnel are busy providing clinical care. Due to multiple caregivers, the person entering the data is generally not present for the entire episode of care.
Further, the fact that the data in ObstetriX are collected to a large extent by 16 weeks gestation (supplemented with data from the labour and birth), while hospital data are collected during the admission for the birth and any other admissions during pregnancy, may affect the quality of ObstetriX as a reference standard, as issues may arise or resolve between early pregnancy and birth. This may have resulted in under-enumeration of true cases in ObstetriX and underestimation of PPV. For example, a woman with chronic hypertension may be first diagnosed after the booking visit (since the definition of chronic hypertension in pregnancy is raised blood pressure before 20 weeks). In this case it will be missed at the booking visit due to being undiagnosed, and as a non-pregnancy-induced condition the likelihood that it would be recorded in ObstetriX at the time of birth is low. This is unlikely to be the case for other conditions, however, including gestational hypertension, as pregnancy-induced conditions would be recorded in the data collected at birth. Nevertheless, where linked data are available, drawing on both data sources to identify cases is probably the best approach, particularly where a condition may be expected to arise late in pregnancy.
Data were available from two hospitals only. However, the hospitals represent different ethnic and socioeconomic compositions, and the similarities of reporting between hospitals are encouraging.
Our findings suggest that coded hospital data are a reliable source of information for gestational diabetes, pre-existing diabetes and all types of hypertension, with the exception of chronic hypertension. Chronic hypertension was reported moderately well, and reliability would be improved by using a grouped category for any hypertension. As thyroid conditions and morbid obesity were poorly reported, coded hospital data should be used with caution for these conditions and if possible, other sources of data should be sought. While there may be local idiosyncrasies in coding and reporting, NSW hospital data are coded following international coding standards, with ICD-10 widely used worldwide [15, 38], and previous studies have shown similar sensitivities between NSW hospital data and data from elsewhere . We therefore consider these findings to be reasonably generalizable to other settings.
The authors thank the New South Wales Local Health Districts and the hospitals for providing access to the data. We thank the personnel who linked and checked the records. This work was funded by the Sydney Medical School Kick Start Grant Program and the Prevention Research Support Program, funded by the New South Wales Ministry of Health.
|ICD||International Classification of Diseases|
|BMI||body mass index|
|PPV||positive predictive value|
|NPV||negative predictive value|
Cheng H-T, Wang Y-C, Lo H-C, Su L-T, Lin C-H, Sung F-C, Hsieh, C-H. Trauma during pregnancy: a populationbased analysis of maternal outcome. World J Surgery. 2012;36(12):2767–75. 10.1007/s00268-012-1750-6Trauma during pregnancy: a populationbased analysis of maternal outcome. World J Surgery. 2012; 36(12):2767-75. https://doi.org/10.1007/s00268-012-1750-6
Shaheen AAM, Myers RP. The outcomes of pregnancy in patients with cirrhosis: a population-based study. Liver Int. 2010;30(2):275–83. 10.1111/j.1478-3231.2009.02153.xThe outcomes of pregnancy in patients with cirrhosis: a population-based study. Liver Int. 2010; 30(2):275-83. https://doi.org/10.1111/j.1478-3231.2009.02153.x
Nair M, Kurinczuk JJ, Knight M. Establishing a national maternal morbidity outcome indicator in England: a population-based study using routine hospital data. PLOS One. 2016;11(4):e0153370. 10.1371/journal.pone.0153370Establishing a national maternal morbidity outcome indicator in England: a population-based study using routine hospital data. PLOS One. 2016; 11(4):e0153370. https://doi.org/10.1371/journal.pone.0153370
Lee Y, Roberts CL, Dobbins T, Stavrou E, Black K, Morris J, Young J. Incidence and outcomes of pregnancyassociated cancer in Australia, 1994–2008: a populationbased linkage study. BJOG. 2012;119(13):1572–82. 10.1111/j.1471-0528.2012.03475.xIncidence and outcomes of pregnancyassociated cancer in Australia, 1994–2008: a populationbased linkage study. BJOG. 2012; 119(13):1572-82. https://doi.org/10.1111/j.1471-0528.2012.03475.x
O’Reilly J, Busse R, Hakkinen U, Or Z, Street A, Wiley M. Praying for hospital care: the experience with implementing activity-based funding in five European countries. Health Econ Pol Law. 2012;7:73. 10.1017/S1744133111000314Praying for hospital care: the experience with implementing activity-based funding in five European countries. Health Econ Pol Law. 2012; 7:73. https://doi.org/10.1017/S1744133111000314
Mehrabadi A, Liu S, Bartholomew S, Hutcheon JA, Magee LA, Kramer MS, Liston RM, Joseph KS, Canadian Perinatal Surveillance System Public Health Agency of Canada. Hypertensive disorders of pregnancy and the recent increase in obstetric acute renal failure in Canada: population based retrospective cohort study. BMJ. 2014;349:g4731. 10.1136/bmj.g4731Canadian Perinatal Surveillance System Public Health Agency of Canada. Hypertensive disorders of pregnancy and the recent increase in obstetric acute renal failure in Canada: population based retrospective cohort study. BMJ. 2014; 349:g4731. https://doi.org/10.1136/bmj.g4731
Hapo Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med. 2008;358(19):1991–2002. 10.1056/NEJMoa0707943Hapo Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med. 2008; 358(19):1991-2002. https://doi.org/10.1056/NEJMoa0707943
Alemu A, Terefe B, Abebe M, Biadgo B. Thyroid hormone dysfunction during pregnancy: a review. Int J Reprod Biomed. 2016;14(11):677–86. URL: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5153572/Thyroid hormone dysfunction during pregnancy: a review. Int J Reprod Biomed. 2016; 14(11):677-86.
Fadl HE, Östlund I, Magnuson A, Hanson US. Maternal and neonatal outcomes and time trends of gestational diabetes mellitus in Sweden from 1991 to 2003. Diabet Med. 2010;27(4):436–41. 10.1111/j.1464-5491.2010.02978.xMaternal and neonatal outcomes and time trends of gestational diabetes mellitus in Sweden from 1991 to 2003. Diabet Med. 2010; 27(4):436-41. https://doi.org/10.1111/j.1464-5491.2010.02978.x
Ray J, Vermeulen M, Shapiro J, Kenshole A. Maternal and neonatal outcomes in pregestational and gestational diabetes mellitus, and the influence of maternal obesity and weight gain: the DEPOSIT* study. QJM. 2001;94(7):347–56. 10.1093/qjmed/94.7.347Maternal and neonatal outcomes in pregestational and gestational diabetes mellitus, and the influence of maternal obesity and weight gain: the DEPOSIT* study. QJM. 2001; 94(7):347-56. https://doi.org/10.1093/qjmed/94.7.347
Roberts CL, Ford JB, Henderson-Smart DJ, Algert CS, Morris JM. Hypertensive disorders in pregnancy: a population-based study. MJA. 2005;182(7):332–5. 10.5694/j.1326-5377.2005.tb06730.xHypertensive disorders in pregnancy: a population-based study. MJA. 2005; 182(7):332-5. https://doi.org/10.5694/j.1326-5377.2005.tb06730.x
Bell JC, Ford JB, Cameron CA, Roberts CL. The accuracy of population health data for monitoring trends and outcomes among women with diabetes in pregnancy. Diabet Res Clin. 2008;81(1):105–9. 10.1016/j.diabres.2008.03.001The accuracy of population health data for monitoring trends and outcomes among women with diabetes in pregnancy. Diabet Res Clin. 2008; 81(1):105-9. https://doi.org/10.1016/j.diabres.2008.03.001
Roberts CL, Bell JC, Ford JB, Hadfield RM, Algert CS, Morris JM. The accuracy of reporting of the hypertensive disorders of pregnancy in population health data. Hyperten Pregnancy. 2008;27(3):285–97. 10.1080/10641950701826695The accuracy of reporting of the hypertensive disorders of pregnancy in population health data. Hyperten Pregnancy. 2008; 27(3):285-97. https://doi.org/10.1080/10641950701826695
Taylor LK, Travis S, Pym M, Olive E, Henderson-Smart DJ. How useful are hospital morbidity data for monitoring conditions occurring in the perinatal period? Aust N Z J Obstet Gynaecol. 2005;45(1):36–41. 10.1111/j.1479-828x.2005.00339.xHow useful are hospital morbidity data for monitoring conditions occurring in the perinatal period?. Aust N Z J Obstet Gynaecol. 2005; 45(1):36-41. https://doi.org/10.1111/j.1479-828x.2005.00339.x
Lain SJ, Hadfield RM, Raynes-Greenow CH, Ford JB, Mealing NM, Algert CS, Roberts CL. Quality of data in perinatal population health databases: a systematic review. Med Care. 2012;50(4):e7–e20. 10.1097/mlr.0b013e31821d2b1dQuality of data in perinatal population health databases: a systematic review. Med Care. 2012; 50(4):e7-e20. https://doi.org/10.1097/mlr.0b013e31821d2b1d
Yasmeen S, Romano PS, Schembri ME, Keyzer JM, Gilbert WM. Accuracy of obstetric diagnoses and procedures in hospital discharge data. Am J Obstet Gynecol. 2006;194(4):992–1001. 10.1016/j.ajog.2005.08.058Accuracy of obstetric diagnoses and procedures in hospital discharge data. Am J Obstet Gynecol. 2006; 194(4):992-1001. https://doi.org/10.1016/j.ajog.2005.08.058
Patterson JA, Francis S, Ford JB. Assessing the accuracy of reporting of maternal red blood cell transfusion at birth reported in routinely collected hospital data. Matern Child Health J. 2016;20(9):1878–85. 10.1007/s10995-016-1992-2Assessing the accuracy of reporting of maternal red blood cell transfusion at birth reported in routinely collected hospital data. Matern Child Health J. 2016; 20(9):1878-85. https://doi.org/10.1007/s10995-016-1992-2
Patterson JA, Roberts CL, Taylor LK, Ford JB. Reporting postpartum haemorrhage with transfusion: a comparison of NSW birth and hospital data. NSW Pub Health Bulletin. 2014;24(4):153–8. 10.1071/NB13008Reporting postpartum haemorrhage with transfusion: a comparison of NSW birth and hospital data. NSW Pub Health Bulletin. 2014; 24(4):153-8. https://doi.org/10.1071/NB13008
Aboa-Eboulé C, Mengue D, Benzenine E, Hommel M, Giroud M, Béjot Y, Quantin C. How accurate is the reporting of stroke in hospital discharge data? A pilot validation study using a population-based stroke registry as control. J Neurol. 2013;260(2):605–13. 10.1007/s00415-012-6686-0How accurate is the reporting of stroke in hospital discharge data? A pilot validation study using a population-based stroke registry as control. J Neurol. 2013; 260(2):605-13. https://doi.org/10.1007/s00415-012-6686-0
- Australian Consortium for Classification Development, Australian Coding Standards for the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification, Tenth Edition and the Australian Classification of Health Interventions. Independent Hospital Pricing Authority: Australia; 2017.
Hadfield RM, Lain SJ, Cameron CA, Bell JC, Morris JM, Roberts CL. The prevalence of maternal medical conditions during pregnancy and a validation of their reporting in hospital discharge data. Aust N Z J Obstet Gynaecol. 2008;48(1):78–82. 10.1111/j.1479-828x.2007.00818.xThe prevalence of maternal medical conditions during pregnancy and a validation of their reporting in hospital discharge data. Aust N Z J Obstet Gynaecol. 2008; 48(1):78-82. https://doi.org/10.1111/j.1479-828x.2007.00818.x
Wacholder S, Armstrong B, Hartge P. Validation studies using an alloyed gold standard. Am J Epidemiol. 1993;137(11):1251–8. 10.1093/oxfordjournals.aje.a116627Validation studies using an alloyed gold standard. Am J Epidemiol. 1993; 137(11):1251-8. https://doi.org/10.1093/oxfordjournals.aje.a116627
Reitsma JB, Rutjes AW, Khan KS, Coomarasamy A, Bossuyt PM. A review of solutions for diagnostic accuracy studies with an imperfect or missing reference standard. J Clin Epidemiol. 2009;62(8):797–806. 10.1016/j.jclinepi.2009.02.005A review of solutions for diagnostic accuracy studies with an imperfect or missing reference standard. J Clin Epidemiol. 2009; 62(8):797-806. https://doi.org/10.1016/j.jclinepi.2009.02.005
Fottrell E, Byass P, Berhane Y. Demonstrating the robustness of population surveillance data: implications of error rates on demographic and mortality estimates. BMC Med Res Methodol. 2008;8(1):13. 10.1186/1471-2288-8-13Demonstrating the robustness of population surveillance data: implications of error rates on demographic and mortality estimates. BMC Med Res Methodol. 2008; 8(1):13. https://doi.org/10.1186/1471-2288-8-13
- Centre for Epidemiology and Evidence. NSW Ministry of Health: Sydney; 2019.
Luef BM, Andersen LB, Renäult KM, Nohr EA, Jørgensen JS, Christesen HT. Validation of hospital discharge diagnoses for hypertensive disorders of pregnancy. Acta Obstet Gynecol Scand. 2016;95(11):1288–94. 10.1111/aogs.13006Validation of hospital discharge diagnoses for hypertensive disorders of pregnancy. Acta Obstet Gynecol Scand. 2016; 95(11):1288-94. https://doi.org/10.1111/aogs.13006
Lain SJ, Roberts CL, Hadfield RM, Bell JC, Morris JM. How accurate is the reporting of obstetric haemorrhage in hospital discharge data? A validation study. Aust N Z J Obstet Gynaecol. 2008;48(5):481–4. 10.1111/j.1479-828X.2008.00910.xHow accurate is the reporting of obstetric haemorrhage in hospital discharge data? A validation study. Aust N Z J Obstet Gynaecol. 2008; 48(5):481-4. https://doi.org/10.1111/j.1479-828X.2008.00910.x
Burns EM, Rigby E, Mamidanna R, Bottle A, Aylin P, Ziprin P, Faiz OD. Systematic review of discharge coding accuracy. J Public Health. 2012;34(1):138–48. 10.1093/pubmed/fdr054Systematic review of discharge coding accuracy. J Public Health. 2012; 34(1):138-48. https://doi.org/10.1093/pubmed/fdr054
Forehan S. Thyroid disease in the perinatal period. Aust Fam Physician 2012;41(8):578–81. URL: https://pubmed.ncbi.nlm.nih.gov/23145397/Thyroid disease in the perinatal period. Aust Fam Physician. 2012; 41(8):578-81.
Australian Institute of Health and Welfare. Incidence of gestational diabetes in Australia 2019 [2 August 2020]. Available from: https://www.aihw.gov.au/reports/diabetes/incidence-of-gestational-diabetes-in-australia/contents/changing-trendsAustralian Institute of Health and Welfare. Incidence of gestational diabetes in Australia 2019 [2 August 2020].
- AIHW: Bulletin no. 146. Cat. no. CDK 7. Canberra; 2019.
- Australian Institute of Health and Welfare. AIHW: Canberra; 2017.
- Australian Institute of Health and Welfare. AIHW: Canberra; 2013.
Healthstats NSW. Maternal medical conditions, NSW 2007 to 2018: NSW Government; 2020 [2 August 2020]. Available from: http://www.healthstats.nsw.gov.au/Indicator/mab_mmedi_cat/mab_mmedi_cat_trend?&topic=Mothers and babies&topic1=topic_mab&code=mum mabMaternal medical conditions, NSW 2007 to 2018: NSW Government; 2020 [2 August 2020].
Cleland V, Hume C, Crawford D, Timperio A, Hesketh K, Baur L, Welch N, Salmon J, Ball K. Urban–rural comparison of weight status among women and children living in socioeconomically disadvantaged neighbourhoods. Med J Aust. 2010;192(3):137–40. 10.5694/j.1326-5377.2010.tb03451.xUrban–rural comparison of weight status among women and children living in socioeconomically disadvantaged neighbourhoods. Med J Aust. 2010; 192(3):137-40. https://doi.org/10.5694/j.1326-5377.2010.tb03451.x
Brown J, Kapurubandara S, McGee TM. Confounding effect of ethnic diversity on booking-in body mass index and prevalence of gestational diabetes and hypertensive disorders in pregnant women in western Sydney 1997–2016. Aust N Z J Obstet Gynaecol. 2020;60(3):369–75. 10.1111/ajo.13077Confounding effect of ethnic diversity on booking-in body mass index and prevalence of gestational diabetes and hypertensive disorders in pregnant women in western Sydney 1997–2016. Aust N Z J Obstet Gynaecol. 2020; 60(3):369-75. https://doi.org/10.1111/ajo.13077
Dietz P, Bombard J, Mulready-Ward C, Gauthier J, Sackoff J, Brozicevic P, Gambatese M, Nyland-Funke M, England L, Harrison L, Taylor A. Validation of self-reported maternal and infant health indicators in the Pregnancy Risk Assessment Monitoring System. Matern Child Health J. 2014;18(10):2489–98. 10.1007/s10995-014-1487-yValidation of self-reported maternal and infant health indicators in the Pregnancy Risk Assessment Monitoring System. Matern Child Health J. 2014; 18(10):2489-98. https://doi.org/10.1007/s10995-014-1487-y
Kim JY, Beckwith BA. The coming wave of change: ICD-10. J Pathol Inform. 2010;1:28. 10.4103/2153-3539.74183The coming wave of change: ICD-10. J Pathol Inform. 2010; 1:28. https://doi.org/10.4103/2153-3539.74183
This work is licensed under a Creative Commons Attribution 4.0 International License.