Clinical Inertia in the Diagnosis and Management of Hypertension Following Ambulatory Blood Pressure Monitoring

Abstract

BACKGROUND

Clinical inertia is common after office blood pressure (BP) is high. Little is known about clinical inertia after ambulatory BP monitoring (ABPM).

METHODS

This was an electronic health record-based retrospective cohort study of patients with high office BP (≥140/90 mm Hg) referred for ABPM at a medical center in New York City between 2016 and 2020. Diagnostic inertia was defined as clinicians not newly diagnosing or treating hypertension in patients with high ABPM (i.e., mean awake BP ≥135/85 mm Hg). Therapeutic inertia was defined as clinicians not intensifying treatment for patients with established hypertension after high ABPM. Multilevel modeling was used to assess patient and clinician characteristics associated with inertia.

RESULTS

Among 329 patients without prior hypertension, 144 (44%) had high awake BP. Of these, diagnostic inertia occurred in 45 of 144 (31%). Among 239 patients taking antihypertensive medication, 141 (59%) had high awake BP. Of these, therapeutic inertia occurred in 73 of 141 (52%). In multilevel models, male gender (odds ratio [OR] 2.81, 95% confidence interval [CI] 1.11–7.08), lower awake systolic BP (SBP) (OR 0.73 per 5 mm Hg increase, 95% CI 0.53–1.00), and specialist vs. primary care clinician type (OR 4.57, 95% CI 1.78–11.75) were associated with increased diagnostic inertia. Increasing age (OR 1.16 per 5-year increase, 95% CI 1.00–1.28) and lower awake SBP (OR 0.82 per 5 mm Hg increase, 95% CI 0.66–0.95) were associated with increased therapeutic inertia.

CONCLUSIONS

Diagnostic and therapeutic inertia were common after ABPM, particularly when awake SBP was near the threshold.

Graphical Abstract

Graphical Abstract.

Hypertension affects 120 million US adults and predicts the risk of cardiovascular disease, disability, and death.¹ While hypertension can be reliably controlled with blood pressure (BP) medications and lifestyle changes, only 43% of adults in the United States with hypertension achieve BP control when using thresholds recommended by recent guidelines.² Failure to diagnose hypertension despite evidence that it is present and failure to intensify treatment despite evidence that hypertension is uncontrolled, known as diagnostic and therapeutic inertia, respectively, are major contributors to uncontrolled hypertension.^3–6 Clinical inertia (encompassing both diagnostic and therapeutic inertia) occurs at more than one-third, and in some studies up to 87% of office visits at which BP is above goal when strictly defined in terms of intensifying the antihypertensive medication regimen.^7–9 Variability between BP readings and differences in BP according to the setting contribute to uncertainty about whether to diagnose or intensify therapy for patients with high office BP which, in turn, promotes clinical inertia.¹⁰

To improve the accuracy of BP measurement and the appropriateness of hypertension diagnosis and management, guidelines from the 2017 American College of Cardiology (ACC), American Heart Association (AHA), and 2021 US Preventive Services Task Force recommend that high office BP be confirmed with out-of-office BP testing such as ambulatory BP monitoring (ABPM) before hypertension is diagnosed or medications are titrated.^11,12 ABPM uptake in the United States has been modest thus far, and consequently little is known about how clinicians use ABPM results in the diagnosis and management of hypertension, particularly in the primary care setting.^13,14

In this study, we aimed to assess the prevalence and correlates of diagnostic and therapeutic inertia in patients with high out-of-office BP by ABPM. Because ABPM provides a more accurate and predictive representation of a patient’s BP and is therefore expected to increase clinician confidence in the accuracy of BP readings, we hypothesized that inertia following high out-of-office BP by ABPM would be uncommon.¹⁵ We also explored the prevalence and correlates of overdiagnosis and overtreatment of hypertension following ABPM in patients without high awake BP.

METHODS

Setting

We conducted a retrospective cohort study of patients referred for ABPM either to establish a new hypertension diagnosis or to determine the adequacy of the current hypertension regimen. Patients were identified from a registry of patients referred to Columbia University Irving Medical Center’s ABPM service from January 2016 to February 2020. Patients could be referred by specialist or primary care clinicians and were eligible to receive ABPM irrespective of health insurance status. Patients were billed for ABPM as part of routine care. The study was approved by the Columbia University Irving Medical Center Institutional Review Board with a waiver of informed consent.

Patients

Patients were included in this analysis if they had completed ABPM at Columbia’s ABPM service for diagnosis or management of hypertension with a minimum of 14 awake BP readings. Patients were excluded if ABPM results were unavailable in the electronic health record (EHR), if they were referred from a practice unaffiliated with Columbia as this precluded chart extraction, or if they were referred because of concern for hypotension rather than hypertension. A fourth-year medical student (J.F.) extracted data pertaining to patient demographic characteristics (age, gender, and preferred language), comorbid cardiometabolic conditions (diabetes, chronic kidney disease, coronary artery disease, or prior myocardial infarction, prior stroke or transient ischemic attack, and depression), total number of chronic disease oral medications, and referring clinician type (specialist or primary care).

ABPM protocol

ABPM was conducted using a SpaceLabs 90227 On Trak device with readings taken every 30 minutes for 24 hours including during sleep. Trained clinical personnel placed ABPM devices and explained their proper use. Patients were also asked to keep a sleep diary, which was used to categorize ABPM readings into awake and asleep periods. ABPM results were emailed to referring clinicians and scanned into the EHR. Both email and EHR results included a brief interpretation stating whether ABPM was high or not high. Prior to 2018, reports provided interpretations using an office BP goal of 140/90 mm Hg, as the threshold for high awake BP and specified that this was equivalent to awake BP 135/85 mm Hg.¹⁶ In 2018, after the publication of the 2017 ACC/AHA US BP guidelines, reports were expanded to also include an office BP goal of 130/80 mm Hg, equivalent to awake BP 130/80 mm Hg, as the threshold for a hypertension diagnosis or treatment intensification in patients with treated hypertension (Supplementary Figure S1 online).

Appropriateness of hypertension management outcomes

The medical student (J.F.) and a board-certified internist (I.K.) independently extracted data from the medical record to determine clinical actions by referring clinicians or clinicians from the referring clinician’s practice within 6 months of ABPM completion. As there was no discordance in classifications from an initial set of records classified by both individuals, the remaining classifications were conducted by 1 extractor. Patients were divided into 2 groups depending on whether they had a preexisting diagnosis of hypertension at the time of ABPM referral. Those referred for potential hypertension diagnosis are hereafter referred to as the diagnostic cohort, and those referred to evaluate their existing hypertension regimen are called the therapeutic cohort.

For patients in both cohorts, we classified the ABPM result as high if average awake systolic BP (SBP) ≥135 mm Hg or diastolic BP ≥85 mm Hg. We chose this threshold as it corresponded to an office BP threshold of 140/90 mm Hg as was recommended by US guidelines prior to 2017 and since the newer threshold of 130/80 mm Hg recommended by the 2017 ACC/AHA BP guidelines was not universally accepted during the study period. We classified ABPM as not high if the average awake SBP <130 mm Hg and diastolic BP <80 mm Hg. Patients with average awake SBP from 130 to 134 mm Hg and average awake diastolic BP from 80 to 84 mm Hg were excluded from primary analyses given the aforementioned change in US BP guidelines during the study period that led to ambiguity in whether BP readings in this range were above or below recommended thresholds. We then assessed the clinician’s action at the subsequent office visit as well as any other chart documentation or orders relevant to hypertension within a 6-month time period. Clinician action was coded as diagnosed or did not diagnose hypertension for the diagnostic cohort and intensified or did not intensify treatment for the therapeutic cohort. The primary method for determining whether hypertension was diagnosed was the addition of hypertension to the problem list and/or documentation of lifestyle counseling and/or BP medication prescribing in the medical note. Intensifying the BP medication regimen or non-pharmacologic interventions such as lifestyle modification and interventions to improve adherence to the existing regimen (for the therapeutic cohort) were counted as treatment intensifications. If there was no documentation from the referring clinician or a clinician of the same specialty within the same practice of diagnosing hypertension or intensifying treatment within the 6-month time period, then the clinical action was coded as hypertension not diagnosed or treatment not intensified, as appropriate.

Diagnostic inertia was classified as present when a patient in the diagnostic cohort had high awake BP and did not receive a new hypertension diagnosis. Therapeutic inertia was classified as present when a patient in the therapeutic cohort had high awake BP and did not have treatment intensified. Overdiagnosis was classified as present when a patient in the diagnostic cohort did not have high awake BP but received a hypertension diagnosis. Overtreatment was classified as present when a patient in the therapeutic cohort did not have high awake BP and had treatment intensified.

In cases of discordance between ABPM result and clinician action, the chart abstractor inferred the reason from the visit note and placed the reason into one of several categories of discordance (e.g., “clinician unaware of ABPM results,” “clinician could not locate ABPM results in EHR,” and “office BP normal at follow-up visit”). When the extractor was unsure of how to classify the reason, consensus was achieved by discussion with the other chart extractor.

Statistical analysis

Descriptive statistics were used to describe the patient sample and the prevalence of diagnostic/therapeutic inertia and overdiagnosis/overtreatment. Descriptive statistics were also used to describe the prevalence of initiating or intensifying antihypertensive medication in patients with intermediate mean awake ABPM readings who were otherwise excluded from analyses. Sensitivity analyses were conducted in which diagnostic and therapeutic inertia were based on 24-hour BP in place of awake BP; 24-hour BP was defined as high if average 24-hour SBP ≥130 mm Hg and average 24-hour diastolic BP ≥80 mm Hg and as not high if average 24-hour SBP <130 mm Hg and average 24-hour diastolic BP <80 mm Hg. Chi-squared and t-tests were used to compare patients with and without inertia. Generalized estimating equations were used to determine the association between patient demographic characteristics (age and sex), chronic conditions (depression and diabetes), out-of-office BP (deviation of average awake systolic and diastolic BP from thresholds), and clinician type (specialist vs. primary care) with diagnostic and therapeutic inertia, accounting for clustering of patients within referring clinicians. Variables were selected based on a review of prior literature on factors associated with clinical inertia.^6,7 A P value <0.05 was used to indicate statistical significance.

RESULTS

Patients referred to ABPM for diagnosis (diagnostic cohort)

Three hundred twenty-nine patients were referred by 136 different clinicians for ABPM to confirm a hypertension diagnosis in patients with high office BP. Overall, the mean age of patients was 52 years, and 74% were female (Table 1). ABPM was high in 144 of 329 (44%) of such patients, and of those patients, 45 of 144 (31%) experienced diagnostic inertia (Figure 1).

Table 1.

Characteristics of patients referred for ambulatory blood pressure monitoring

Characteristic	No prior diagnosis of hypertension and high ABPMa (N = 145)	No prior diagnosis of hypertension and not high ABPMb (N = 114)	Treated for hypertension and high ABPMa (N = 141)	Treated for hypertension and not high ABPMb (N = 68)
Age in years, mean (SD)	53.3 (14.5)	52.1 (17.6)	62.6 (16.1)	64.2 (12.0)
Gender
Female	105 (72.4%)	87 (76.3%)	88 (62.4%)	48 (70.6)
Male	40 (27.6%)	27 (23.7%)	53 (37.6%)	20 (29.4)
Preferred language
English	77 (53.1%)	48 (42.1%)	81 (57.4%)	38 (55.9)
Spanish	65 (44.8%)	61 (53.5%)	59 (41.8%)	29 (42.6)
Other	3 (2.1%)	5 (4.4%)	1 (0.7%)	1 (1.5)
Mean (SD) awake SBP in mm Hg	141.1 (8.8)	120.0 (5.3)	144.8 (12.1)	119.7 (5.8)
Mean (SD) awake DBP in mm Hg	85.3 (8.0)	72.7 (4.9)	81.2 (11.0)	71.0 (5.4)
Diabetes	9 (6.2%)	9 (7.9%)	35 (24.8%)	20 (29.4%)
Chronic kidney disease	6 (4.1%)	1 (2.6%)	15 (10.6%)	5 (7.4%)
CVD/prior MI	2 (1.4%)	1 (1.8%)	20 (14.2%)	10 (14.7%)
Stroke/TIA	4 (2.8%)	3 (2.6%)	21 (14.9%)	5 (7.4%)
Depression	36 (24.8%)	23 (20.2%)	26 (18.4%)	9 (13.2%)
Mean (SD) number of chronic medications	4.3 (4.1)	4.1 (3.5)	7.1 (4.4)	7.4 (3.9)
Referring clinician type
Primary care	126 (86.9%)	102 (89.5%)	82 (58.2%)	47 (69.1%)
Specialist	19 (13.1%)	11 (9.6%)	59 (41.8%)	21 (30.9%)
Missing	0 (0%)	1 (0.9%)	0 (0%)	0 (0%)

Abbreviations: ABPM, ambulatory blood pressure monitoring; CVD, cardiovascular disease; DBP, diastolic blood pressure; MI, myocardial infarction; SBP, systolic blood pressure; TIA, transient ischemic attack.

^aHigh ABPM defined as average awake BP ≥135/85 mm Hg.

^bNot high ABPM defined as average awake BP <130/80 mm Hg.

Figure 1.

Clinician management after ABPM in patients with and without prior diagnosis of hypertension and high office blood pressure. Abbreviation: ABPM, ambulatory blood pressure monitoring.

Approximately half of diagnostic inertia (23 of 45) occurred in patients who had no follow-up visits with the referring clinician (or a clinician from the same practice) within 6 months, and half (22 of 45) occurred in patients who had at least 1 follow-up visit within this period. The most common reasons for diagnostic inertia at follow-up visits were lack of clinician awareness of ABPM results and borderline ABPM results (i.e., average awake SBP 135–139 mm Hg and average awake diastolic BP from 85 to 59 mm Hg; Figure 2). Among those who were appropriately diagnosed with hypertension based on high ABPM (99 of 144; 69%), prescribing a new antihypertensive medication was the most common type of related clinical action (76 of 99 patients; 77%).

Figure 2.

Reasons for diagnostic and therapeutic inertia in patients with high ABPM (i.e., average awake BP ≥135/85 mm Hg). Abbreviations: ABPM, ambulatory blood pressure monitoring; BP, blood pressure.

In multilevel analyses assessing correlates of diagnostic inertia in patients with high ABPM (Table 2), women vs. men (adjusted odds ratio [aOR] 0.36, 95% confidence interval [CI] 0.14–0.90) and higher awake SBP (greater positive deviation from the 135 mm Hg threshold) (OR 0.73 per 5 mm Hg increase, 95% CI 0.53–1.00) were associated with lower odds of diagnostic inertia whereas specialist vs. primary care clinician type was associated with greater odds of diagnostic inertia (aOR 4.57, 95% CI 1.78–11.75).

Table 2.

Correlates of diagnostic inertia after high awake ambulatory blood pressure in patients without a prior history of hypertension

Characteristic	OR	95% CI	P value
Patient level
Age (per 1-yr increase)	1.01	0.97–1.04	0.70
Women (vs. men)	0.36	0.14–0.90	0.03
Diabetes (vs. no diabetes)	0.59	0.09–4.06	0.59
Depression (vs. no depression)	0.39	0.11–1.37	0.14
SBP deviation (per 1 mm Hg increase)	0.94	0.88–1.00	0.04
DBP deviation (per 1 mm Hg increase	0.99	0.92–1.05	0.68
Clinician level
Specialist (vs. primary care)	4.57	1.78–11.75	0.002

Abbreviations: CI, confidence interval; DBP, diastolic blood pressure; OR, odds ratio; SBP, systolic blood pressure. Bold indicates characteristics associated with diagnostic inertia after high awake SBP at P < .05.

ABPM was not high in 114 of 329 (35%) patients referred with high office BP, which corresponds to the proportion with white-coat hypertension. Of those patients, 3 of 114 (3%) experienced overdiagnosis (Figure 1). Reasons for overdiagnosis were high office BP at visits following referral for ABPM (N = 2) and isolated nocturnal hypertension (N = 1).

In a sensitivity analysis in which ABPM was classified according to 24-hour BP, ABPM was high in 155 of 265 patients (59%), and of those patients, 53 of 155 patients (34%) experienced diagnostic inertia. In a secondary analysis describing actions taken for patients with intermediate awake BP readings (i.e., awake SBP 130–135 mm Hg and awake BP 80–85 mm Hg), hypertension was diagnosed in 15 of 71 (21%) patients in this group.

Patients referred to ABPM for treatment decision-making (therapeutic cohort)

Two hundred thirty-nine patients were referred by 112 different clinicians for ABPM to evaluate out-of-office BP in patients with treated hypertension. The mean age was 63 years, and 64% of patients were female (Table 1). ABPM was high in 140 of 239 (59%) cases, and 73 of 140 (52%) of these patients experienced therapeutic inertia (Figure 1). As with the diagnostic cohort, approximately half of therapeutic inertia occurred among patients with no follow-up visits within 6 months and half with at least 1 follow-up visit. The most common reasons for therapeutic inertia in those with follow-up visits were a lack of clinician awareness of ABPM results and borderline ABPM results (Figure 2). Among those with appropriate treatment intensification for high ABPM, the addition of a new antihypertensive medication was the most common clinical action followed by an increased dose of a current antihypertensive medication.

In multilevel analyses assessing correlates of therapeutic inertia in patients with high ABPM, higher age (aOR 1.16 per 5-year increase, 95% CI 1.00–1.28), and smaller deviation of awake SBP from the threshold (aOR 0.82 per 5 mm Hg increase, 95% CI 0.66–0.95) were associated with higher odds of therapeutic inertia (Table 3).

Table 3.

Correlates of therapeutic inertia after high awake ambulatory blood pressure in patients with a prior history of hypertension

Characteristic	OR	95% CI	P value
Patient level
Age (per 1-yr increase)	1.03	1.00–1.05	0.04
Women (vs. men)	1.27	0.61–2.61	0.52
Diabetes (vs. no diabetes)	2.11	0.78–5.71	0.14
Depression (vs. no depression)	0.84	0.33–2.16	0.72
Average awake SBP (per 1 mm Hg increase)	0.96	0.92–0.99	0.01
Average awake DBP (per 1 mm Hg increase)	1.00	0.96–1.03	0.91
Clinician level
Specialist (vs. primary care)	1.61	0.74–3.51	0.23

Abbreviations: CI, confidence interval; DBP, diastolic blood pressure; OR, odds ratio; SBP, systolic blood pressure. Bold indicates characteristics associated with therapeutic inertia after high awake SBP at P < .05.

ABPM was not high in 68 of 239 (28%) patients with treated hypertension and high office BP (i.e., white-coat effect), and 4 of 68 (6%) of these patients experienced overtreatment (Figure 2). Reasons for overtreatment were persistently high office BP at subsequent office visits and high patient-reported home BP readings (N = 2) or high home BP readings, alone (N = 2).

In a sensitivity analysis in which ABPM was classified according to 24-hour BP, ABPM was high in 143 of 199 patients (72%), and 76 of 143 patients (53%) experienced therapeutic inertia. In a secondary analysis describing actions taken for patients with intermediate awake BP readings (i.e., awake SBP 130–135 mm Hg and awake BP 80–85 mm Hg), treatment was intensified in 5 of 31 (16%) patients in this group.

DISCUSSION

Among patients with high ABPM, approximately one-third in the diagnostic cohort and half in the therapeutic cohort experienced clinical inertia. The prevalences of diagnostic and therapeutic inertia were similar regardless of whether classifications were based on average awake SBP or 24-hour BP. The prevalence of diagnostic inertia was consistent with that of Hall et al., who found that diagnostic inertia occurred in 31% of patients with high ABPM in one of the only other studies examining inertia after ABPM.¹⁷ These estimates are lower than those found in some studies assessing clinical inertia based on office BP alone,^7,9 suggesting that ABPM may help reduce inertia.

The most common reasons for diagnostic and therapeutic inertia, present in about half of cases, were lack of follow-up within 6 months of testing and lack of clinician awareness that ABPM results were available in the EHR. These findings were consistent with the lack of timely follow-up that commonly occurs after other types of referrals and diagnostic tests in ambulatory care settings.¹⁸ Lack of timely follow-up has also been found in other samples of patients with hypertension and has been a rationale for promoting team-based care for hypertension.^19,20 Excluding patients with high ABPM who had no follow-up after testing would yield more modest diagnostic and therapeutic inertia percentages of 20% and 37%, respectively.

The higher prevalence of therapeutic inertia relative to diagnostic inertia in both our study and prior work suggests that clinicians may be more willing to intensify treatment for an existing diagnosis than to make a new diagnosis or that patients may resist treatment intensification more than initiation. Interestingly, among the small group of patients with intermediate awake BP readings that were otherwise excluded from analyses, diagnosing hypertension or intensifying hypertension treatment was even less common.

Similar to correlates of inertia after office BP, another common contributor to diagnostic and therapeutic inertia was having modestly elevated ABPM results that were close to the high out-of-office BP threshold.^10,21 Clinicians who are not hypertension specialists may be unaware that an out-of-office BP threshold of 135/85 mm Hg corresponds to an office BP threshold of 140/90 mm Hg. Of note, ABPM reports in this study showed the guideline-recommended thresholds for ABPM, but clinicians may have missed this component of the report. As uptake of ABPM increases, efforts to enhance the integration of ABPM results into the EHR to ensure closed loops following referrals and to educate clinicians about recommended thresholds for ABPM may help reduce inertia from these causes.^22,23

In the diagnostic cohort, alone, inertia was more common with specialists than primary care clinicians. This finding deserves further study but may be mediated by lower follow-up within 6 months in specialist patients.

For only 4 of the 286 patients with high ABPM across both cohorts, clinicians noted that ABPM results were high but chose not to follow guidelines. This is a promising result given that lack of confidence in the accuracy of office BP readings is a strong driver of inertia when relying on office BP, alone, to diagnose and treat hypertension.¹⁰

Several patient demographic characteristics were associated with clinical inertia. In the diagnostic cohort, women had a lower probability of inertia than men. In the therapeutic cohort, greater age was associated with a greater probability of inertia, perhaps due to greater concerns regarding polypharmacy and less intensive BP goals in older adults. All of these analyses were exploratory in nature and should be replicated in larger samples before conclusions are drawn.

In contrast to diagnostic and therapeutic inertia, overdiagnosis and overtreatment were uncommon when ABPM was not high. In those few cases in which they occurred, persistently high office or home BP readings were documented in the EHR, suggesting clinicians may have been reticent to defer to the ABPM results in this context. The discrepancy between ABPM and home BP occurs in approximately 10% of patients who use both modalities, and there is little guidance for management in this scenario.²⁴ Thus, these 2 cases in which ABPM and home BP readings differed in the current study, although considered overtreatment by our study design, involved clinically reasonable actions by clinicians.

Strengths and limitations

There were several limitations to our findings. The study was conducted at a single academic medical center that had suboptimal integration of ABPM results into the EHR, which may limit its generalizability. That said, a racially and ethnically diverse group of patients were referred from a mix of safety net and faculty practices and suboptimal integration of specialty test results into the EHR is not uncommon in other health systems. In cases of discrepancy between ABPM results and clinician action, clinicians’ reasoning was not always explicitly documented in EHR notes and required coder interpretation. Our classification of therapeutic inertia may have missed instances of lifestyle or adherence counseling that were not documented in the EHR which may have led to an overestimate of the prevalence of therapeutic inertia. The relatively small sample size limited the power to precisely estimate the magnitude of the association of correlates with inertia in regression models, and significance testing did not adjust for multiple comparisons. Accordingly, interpretations of these findings should be made cautiously. Furthermore, a limited number of correlates were assessed. We relied on awake BP and 24-hour BP as indicators of high ABPM; the impact of asleep BP on clinical inertia was not studied. Nevertheless, this represents one of the first studies examining the real-world impact of ABPM on clinician diagnosis and management of hypertension when an ABPM testing service is made available in a US setting.

Diagnostic and therapeutic inertia were common after ABPM, though were possibly lower than after office BP, alone. Key drivers included lack of timely follow-up after ABPM and proximity of average awake BP to thresholds for diagnosing or intensifying treatment. Structured training highlighting correct ABPM interpretation and the importance of timely follow-up may need to be integrated into medical school and internal medicine residency curricula and into professional programs, conferences, or workshops targeted at practicing physicians and other advanced practice providers who order ABPM.²⁵ Future efforts to optimize clinician management after ABPM testing may also require more seamless integration of ABPM results into the EHR with enhanced computerized clinical decision support.²⁶

SUPPLEMENTARY DATA

Supplementary materials are available at American Journal of Hypertension (http://ajh.oxfordjournals.org).

FUNDING

Dr Kronish received support from the National Heart, Lung, and Blood Institute (R01 HL152699). Dr Flatow received support from the National Institute of Diabetes and Digestive and Kidney Diseases (T35DK093430).

CONFLICT OF INTEREST

The authors have no conflicts of interest to report.

DATA AVAILABILITY

Deidentified data available upon request.