Improving Blood Pressure Accuracy in the Outpatient Adolescent Setting.

INTRODUCTION: Hypertension (HTN) is increasing in the pediatric population, and hypertensive children become hypertensive adults. Blood pressure (BP) is often overlooked due to factors including distrust in readings, lack of time in visits, and discomfort prescribing antihypertensive medications. The objective of our multistage, clinically based quality improvement project was to improve BP measurement and HTN diagnosis and intervention in adolescents.
METHODS: Study investigators performed interventions in an adolescent medicine clinic at Children's Hospital Colorado (ages 12-24 years), which included equipment inventory, mapping clinic processes, manual/automated BP training for staff, education of faculty/staff on guidelines, and creation/implementation of updated best practice alerts (BPAs) based on age-appropriate guidelines for stage 1, 2 HTN in patients younger/older 18years.
RESULTS: With equipment updates, medical assistant manual BP certifications, educational sessions for faculty/staff, and creation of a BPA with instruction before the go-live date, confidence in using automated and manual BP measurements increased for faculty/staff. The number of unique patient visits presenting with elevated BPs decreased significantly in the postintervention period reflecting the education/training. Staff used the new order set minimally, but there was an increase in correct diagnoses of elevated BP and laboratory workup.
CONCLUSIONS: Having appropriate equipment while introducing BP guideline education in the ambulatory setting with electronic health record reminders and utility of BPAs can decrease erroneous BP values saving providers and staff encounter time. Real-time alerts can aid in accurate diagnosis rates and improved intervention for youth with elevated BP readings. Providers still inconsistently interact with order sets despite such parameters.

INTRODUCTION

Elevated blood pressure (BP) and hypertension (HTN) are increasing in children and have short- and long-term health consequences. The prevalence of prehypertension in adolescents is approximately 14% in boys and 6% in girls, and 14% of youth with prehypertension will develop HTN within 2 years. Five percent will have sustained HTN.[1-3] Hypertensive youth become hypertensive adults with increased obesity, metabolic syndrome, cardiovascular disease, and diabetes.[4]

BP in children and adolescents is a widely overlooked vital sign due to inaccurate, inconsistent automatic BP readings, no consensus evaluating secondary HTN, and discomfort initiating antihypertensive medications.[5] Additionally, the pediatric HTN criteria are complex, with BP cutoffs varying by sex, age, and height. During this project, pediatric guidelines defined HTN as 3 sequential elevated values with stage 1 HTN between 95th and 99th percentiles and stage 2 HTN > 99th percentile + 5 mm Hg for age, sex, and height for youth younger than 17 years.[6] These parameters require diligent referencing for each patient. Data suggest that <30% of children with elevated BP in an electronic medical record were diagnosed with hypertension, and BP is not recorded for 6% of patients.[2]

Adolescent health providers, following patients into their 20s, may diagnose HTN in pediatric and adult populations. Based upon accepted guidelines, adult patients 18 years and older of age with 3 or more readings of >140/>90 mm Hg meet the criteria for HTN.[7] Secondary causes of HTN are less likely in this age group. As adolescent HTN diagnosis and treatment can be complicated, we proposed a multistage, clinic-based quality improvement (QI) project to improve the diagnosis and evaluation of HTN in adolescents. This project’s global aim was to improve BP measurement standardization and efficiency in the adolescent medicine outpatient setting and identify and decrease barriers to diagnosis and intervention of HTN. Our specific aim was to decrease the number of BP readings per visit utilizing education and proper measurement techniques from a baseline of 83%–95% over 1 years’ time.

METHODS

Setting

The project occurred in the Adolescent Medicine Clinic at Children’s Hospital Colorado (6,000 yearly visits; patients 12–24 years old). Please see the project timeline for further details (Fig. 1). The Institutional Review Board approved this study as a QI study. For the study data collection and BPA, we utilized our electronic health record, Epic (Epic Systems, Verona, Wis.), a cloud-based EHR used in community hospitals, independent practices, and multispecialty hospital groups.

Fig. 1.

Project timeline of all interventions.

Baseline Clinical Assessments

Equipment Inventory

We conducted a systematic inventory and equipment functionality check of each examination and intake room. Investigators assessed BP cuff sets and tubing and the functionality of sphygmomanometer dials. Based on assessments, clinic staff repaired or corrected equipment.

Automatic BP Equipment Validation

We performed validation of DinaMap (GE Healthcare, Chicago) automatic BP monitors by comparing medical assistant (MA) DinaMap readings to a research assistant’s (RA) manual average of 2 readings utilizing a convenience sample of clinic patients. Manual readings were performed 1–2 minutes following the DinaMap reading in the clinic intake room to minimize position changes leading to possible BP elevations. The RA was a control for automatic BP readings after their manual BP readings were precise within 5 mm Hg in systolic/diastolic measurements for >50% of measurements (N = 22) compared with 2 clinical staff members (MD and RN) previously trained in accurate manual BP measurements. This validation methodology follows accepted protocols for automatic BP monitors.[8]

Observed BP Clinic Process

The RA observed intake processes, including vital signs and room timing for a convenience sample of 5–10 patient encounters/MA. She also recorded automatic versus manual BP timing and evaluated each intake process with a checklist (including patient and cuff positioning) to delineate variables affecting BP accuracy (see Appendix A, Supplemental Digital Content 1, http://links.lww.com/PQ9/A258).[8]

Knowledge Assessments

Providers and MAs received knowledge assessments evaluating baseline BP/HTN knowledge and comfort with manual BP measurements (see Appendix B, Supplemental Digital Content 2, http://links.lww.com/PQ9/A259). An anonymous survey assessed provider attitudes and trust levels in BP measurements before/after an educational intervention.

Primary Chart Review

Investigators performed the chart review and data collection from Epic (March–May 2016). The study surveyed visit encounters (all patient visits for patients 12–24 years excluding sick, eating disorder visit types) for BP ≥ 95th percentile for younger than 18 years and stage 1 HTN level (≥140/≥90) or above for patients 18 years and older. Data included BP readings, personal/family history, body mass index (BMI), visit diagnostic codes, relevant laboratory codes, and relevant subspecialty referrals. Investigators de-identified all data used for this project.

Interventions

MA Training and Certification

Study investigators conducted in-service training for automatic and manual techniques, including a hands-on workshop based on established guidelines.[6-8] Each MA participated in a certification process; 2 weeks after initial training, MAs had automated/manual BP checklist monitoring. Study personnel compared MA measurements to a provider’s measurement to confirm measurement accuracy. After stage 1 certification, MAs received individualized recommendations for ideal BP measurement. Then MAs received certification in manual BP monitoring after at least 5 manual BP measurements with BP readings within 10 mm Hg of both systolic/diastolic measurements obtained by the study RA.

Provider Training

Providers (n = 13) received educational training on BP recommendations, diagnostic criteria for pediatric/adult populations, accepted screening for secondary causes of HTN, and treatment recommendations. Case scenarios highlighted essential factors related to HTN diagnosis and treatment. Providers completed a pretest immediately before training and a posttest 2 weeks later.

BPA Creation and Implementation

BPA Development

The research team met with the hospital IT team to discuss an electronic alert. Study investigators chose a clinical best practice alert (BPA) despite reservations about “alert fatigue.” Although providers blame alert fatigue for lack of use, studies suggest that repeated alerts/encounters combined with high patient complexity influenced providers to ignore alerts more than presumed desensitization.[9] Our BPA only appeared in the BPA section of the flowsheet, and pop-up capabilities were disabled. The alert disappeared with interaction or with new BP readings below the alert cutoff.

Four alerts were designed based on guideline diagnostic cutoffs for BP meeting stage 1 or stage 2 HTN for pediatric/adult patients based on the 2004 Fourth Report hypertension guidelines; this project occurred before the 2017 updated diagnostic cutoff alterations.[10] The BPA presented the current visit BP and 2 previous visit BPs with stage 1 or stage 2 HTN also designated. Hyperlinks led to the vitals flowsheet to review or enter repeat BP measurements, and the alert allowed users to indicate they would recheck the BP or utilize the order set. Separate MA/nursing view BPA alerts were created and prompted staff to retake a manual BP before patient rooming. The BPA could fire for MA/nursing alone if a second normal reading was obtained.

BPA-linked Order Set Development

The research team identified laboratory testing and evaluations appropriate for pediatric/adult stage 1 and stage 2 BP BPAs. Order sets included diagnostic code options, follow-up instructions, and automated after-visit summary recommendations such as reducing salt intake.

BPA Go-live Implementation

After training on the BPAs and order sets in the Epic training environment, they were activated for the clinic starting in March 2017. One week before the BPA Epic go-live, the research team introduced the BPA and clinical process to all staff. The study team posted reminder cards with screenshots in clinic triage rooms and provider workspaces. Initial issues included a lack of rotating trainee access to BPA alerts, which delayed our secondary chart review by only 2 weeks. Data collection occurred April–July 2017.

Postintervention Assessments

Quizzes (multiple choice and short-answer questions) were administered to MAs and providers at 3 time points to assess BP measurement and management knowledge. The pretest was performed immediately before the educational sessions. The first posttest was administered 2 weeks after educational interventions, and the second posttest was administered 3 months later to assess retained knowledge.

Postintervention Chart Review

Investigators collected average BP measurements and the number of BP taken per visit at 3 intervals by chart review: preeducation training, posteducation training, and postelectronic health record BPA installation. Investigators conducted analyses with SPSS V.24 (P < 0.05). Changes in systolic and diastolic BP measurements were also tracked by statistical process control charts. The preintervention chart review and analysis occurred from March to May 2016; the posteducational/BPA chart review occurred April–July 2017.

RESULTS

Clinic Flow and Inventory

The study RA observed baseline clinic flow from time in the waiting room, vital signs, and time until provider-initiated visit (see Figs. 1 and 2, Supplemental Digital Content 3 and 4, which display preintervention and proposed state process maps, http://links.lww.com/PQ9/A262 and http://links.lww.com/PQ9/A263). Examining the 13 rooms, we found multiple errors with missing cuffs, duplicate cuffs, malfunctioning sphygmomanometer dials, and air leaks in tubing and/or cuff bladders. The 3 intake rooms had only DinaMap automatic machines and lacked 1–3 of 5 cuff sizes (child, small adult, adult, large adult, and thigh) with missing labeling.

BP Measurements Preintervention

The most prominent error was not having patients rest in a seated position for 3–5 minutes before BP measurements. Correct cuff size was only utilized 33% of the time, and patients or MA talked during measurement ~30% of the time. The RA also noted there were no chairs with armrests for support during BP measurements.

Provider and Staff Baseline BP Perspectives

MAs described a lack of trust in DinaMap readings with frequent machine malfunctions (100% felt malfunctioned sometimes/almost every time). MAs performed automated BP (80%) while providers reported they “sometimes” do manual BPs (n = 12, 92.3%). All providers (n = 13) reported taking BP manually if previous abnormal results occur. Only 5 providers (38.5%) agreed they trusted DinaMap readings; providers felt “the DinaMap is often wrong,” and they used manual BP “because it is more accurate.”

BP Accuracy Posttraining

Manual BP certification for MA occurred by comparing manual BPs to the RA’s measurements with ≤10 mm Hg difference for 80% of readings. There was >80% adherence to patient positioning and cuff location after training (see Table 1, Supplemental Digital Content 5, http://links.lww.com/PQ9/A260).

Audits from the EHR for systolic and diastolic BP averages occurred for three differing 3-month periods: pre-QI project, posteducation training, and post-EHR BPA. Values in each stage demonstrate statistically significant decreases in average systolic and diastolic BP measurements over this 9-month project (P < 0.001) (Fig. 2). Systolic BPs decreased from 122 to 114 post-BPA, whereas diastolic BPs decreased from 67 to 62. Additionally, the number of BP measurements/visit decreased. Visits with only 1 BP measurement increased from 83.3% before any intervention to 93.4% (P < 0.001) after education and BPA initiation (Fig. 3).

Fig. 2.

Systolic and diastolic BP decreases over time; P < 0.001.

Fig. 3.

Number of BPs taken per patient over time; P < 0.001. Orange bar = 1 BP measurement; blue bar = 2 BP measurements; green bar = 3 BP measurements; red bar = 4 BP measurements.

Provider and MA Pretest Results

MAs scored slightly higher than providers on the 9-item techniques pretest (64.4% v. 62.6%), as shown in (see Table 2, Supplemental Digital Content 6, http://links.lww.com/PQ9/A261). Both MA and providers scored relatively low on knowing the time for patient rest before taking a BP (40% and 45%, respectively), whereas MA had higher knowledge of what arm to repeat the BP in if initially high (80% versus 36%).

Intervention Posttests

Posttest 1: 2 Weeks after Training

All providers (n = 13) reported taking manual BP measurement to some extent, and trust in automatic BPs increased at posttest 1, with 61% (n = 8) agreeing that they trust DinaMap readings. Confidence in taking manual BPs also increased as providers were “very confident” in their ability to take a manual BP, increasing from 0 to 8 providers (61.5%).

Posttest 2: 3 Months after Training

The average score for posttest 2 was 82.91% for providers. Nearly, 85% (n = 11) of providers used manual BP measurements, whereas 7 providers (53.8%) stated that they never use DinaMap machines. All providers reported taking manual repeat BP measurements and felt confident/very confident. Twelve providers (92%) knew pediatric BP guidelines somewhat well/well/very well, whereas 85% (n = 11) knew adult guidelines somewhat well/well.

Overall Pretest–Posttest Trends

At the pretest, MAs scored higher than providers (72.5% versus 62.60%), though providers improved their scores, outscoring MA averages at both posttraining time points (64.4% versus 71.94% and 69.44% versus 82.91%, respectively). MAs and providers reporting manual BP measurement sometimes/almost every time/every time went from 77.8% to 93.8%; 87.5% (from 72%) used manual BP for repeat readings. After the training, staff trusted DinaMap readings (agree/strongly agree) >50% of events.

Preintervention Chart Review

Investigators gathered data for all preventive and follow-up visits (n = 935) between March 1, 2016 and May 31, 2016. There were 97 patient visits (10.4%) with an elevated BP > 95th percentile (Table 1). Seventy-eight patients (80.4%) had a systolic BP consistent with stage 1 HTN and 6 were consistent with stage 2 HTN (6.2%) (Table 2). Nineteen patients (19.6%) had a diastolic BP corresponding with stage 1 HTN, and 1 patient had a diastolic BP categorized as stage 2 HTN. Only 19 patients (19.6%) with elevated BP readings received a diagnosis of elevated BP or HTN (Table 2). The pre-existing hospital BPA only fired for patients with BP > 99th percentile, which missed stage 1 HTN at 95% and stage 2 HTN defined as 99% +5 mm Hg, and patients younger than 17 years.

Table 1.

Preintervention and PostIntervention Chart Review Data

	Preintervention	Postintervention
Total, N	935	461
Patients with high BP	97 (10.4%)	31 (6.7%)
Gender
Female	63 (64.9%)	19 (61.3%)
Male	34 (35.1%)	12 (38.7%)
Age
12 and under	5 (5.2%)	2 (6.5%)
13–14 y old	24 (24.7%)	5 (16.1%)
15–16 y old	30 (30.9%)	12 (38.7%)
17–18 y old	35 (36.1%)	9 (29.0%)
19–20 y old	1 (1.0%)	3 (9.7%)
21–22 y old	2 (2.1%)	0.00%
Ethnicity
Not Hispanic or Latino	57 (58.8%)	21 (67.7%)
Hispanic or Latino	37 (38.1%)	10 (32.3%)
Unknown/not reported	3 (3.1%)	0
Race
White	36 (37.1%)	11 (35.5%)
Other	30 (30.9%)	9 (29.0%)
Black/African American	21 (21.6%)	8 (25.8%)
More than one race	5 (5.2%)	2 (6.5%)
American Indian/Alaska Native	2 (2.1%)	0
Unknown/not reported	3 (3.1%)	1 (3.2%)
Relevant diagnoses
HTN	4 (4.1%)	4 (12.9%)
Elevated BP	15 (15.5%)	7 (22.5%)
Obesity dx	8 (8.2%)	5 (16.1%)
Smoking	0	1 (3.2%)
Alcohol	0	1 (3.2%)
Snoring	5 (2.1%)	1 (3.2%)
Family history of HTN	48 (49.5%)	18 (58.1%)
Smoking status
Never smoker	70 (75.2%)	23 (74.2%)
Passive smoke exposure	12 (12.4%)	3 (9.7%)
Current everyday smoker	5 (5.2%)	1 (3.2%)
Former smoker	5 (5.2%)	2 (6.5%)
Missing	3 (3.1%)	1 (3.2%)
Current some day smoker	2 (2.1%)	0
Laboratories done at visit
Total patients receiving any laboratory	38 (39.2%)	21 (67.7%)
Urinalysis/urine dip	3 (3.1%)	7 (22.5%)
Electrolytes	0 (0.0%)	0
BUN/Cr	0 (0.0%)	0
CBC	13 (13.4%)	7 (22.6%)
Lipid panel	27 (27.8%)	13 (41.9%)
A1C/fasting glucose/glucose serum	22 (22.7%)	13 (41.9%)
Thyroid stim hormone	8 (8.2%)	4 (12.9%)
GPT/ALT	11 (11.3%)	6 (19.4%)
GOT/AST	9 (9.3%)	3 (9.7%)
Hepatic function panel	1 (1.0%)	2 (6.5%)
Basic/complete metabolic panel	7 (7.2%)	1 (3.2%)
Renal function panel	4 (4.1%)	3 (9.7%)
Cortisol	0	1 (3.2%)
Imaging/Procedures
Total (any imaging/procedures done)	5 (5.2%)	2 (6.4%)
ECG	2 (2.1%)	1 (3.2%)
Renal ultrasound	3 (3.1%)	1 (3.2%)
Referral
Total (any referral made)	7 (7.2%)	2 (6.4%)
Lifestyle medicine	5 (5.2%)	1 (3.2%)
Cardiology	2 (2.1%)	0
Nephrology	0 (0.0%)	1 (3.2%)
Visit provider gender
Female	69 (71.1%)	23 (74.2%)
Male	28 (28.9%)	8 (25.1%)
Attending provider gender
Female	59 (60.8%)	17 (54.8%)
Male	38 (39.2%)	14 (45.2%)
Trainee involved
Yes	61 (62.9%)	19 (61.3%)
No	36 (37.1%)	12 (38.7%)
Trainee type (n = 57)
Resident
First year	29 (29.9%)	12 (38.7%)
Second year	0 (0.0%)	0 (0.0%)
Third year	1 (1.0%)	0
Fellow
First year	31 (32.0%)	4 (12.9%)
second year	0 (0.0%)	3 (9.7%)
Third year	0 (0.0%)	0 (0.0%)
Trainee to attending gender (n = 57)
Female, female	35 (36.1%)	10 (32.3%)
Female, male	20 (20.6%)	6 (19.4%)
Male, male	1 (1.0%)	3 (9.7%)
Male, female	5 (5.2%)	0

ALT, alanine transaminase; AST, aspartate aminotransferase; BUN/Cr, blood urea nitrogen/creatinine ratio; CBC, complete blood count; dx, diagnosis; ECG, electrocardiogram; GPT, glutamic pyruvic transaminase; GOT, glutamic oxaloacetic transaminase.

Table 2.

Patients Meeting Criteria for Elevated BP or HTN Criteria

	Pre-QI Project			Post-QI Project
Patients with High Blood Pressure	n = 97			n = 31
Mean Age	16.95; P < 0.008*			16; P < 0.504
BMI	31.7; P < 0.404			30.7; P < 0.242
Hypertension/Elevated BP Diagnosis
	Appropriate Diagnosis	No Diagnosis	Sig	Appropriate Diagnosis	No Diagnosis	Sig
Patient gender
Female	11 (57.9%)	52 (66.7%)	0.472	4 (40%)	15 (71.4%)	0.093
Male	8 (42.1%)	26 (33.3%)		6 (60%)	6 (28.6%)
Patient age (y)
Adolescent (<18)	13 (68.4%)	76 (97.4%)	0.001*	9 (90%)	19 (90.5%)	0.967
Adult (18+)	6 (31.6%)	2 (2.6%)		1 (10%)	2 (9.5%)
Ethnicity
Not Hispanic or Latino	10 (52.6%)	47 (60.3%)	0.494	8 (80%)	13 (61.9%)	0.314
Hispanic or Latino	9 (47.4%)	28 (35.9%)		2 (20%)	8 (38.1%)
Unknown/not reported	0	3 (3.8%)		0	0
Race
White	4 (21%)	32 (41.0%)	0.058ⱡ	3 (30%)	8 (38.1%)	0.079
Other	11 (57.9%)	19 (24.4%)		1 (10%)	8 (38.1%)
Black/African American	3 (15.8%)	18 (23.1%)		3 (30%)	5 (23.8%)
More than one race	0	5 (6.4%)		2 (20%)	0
American Indian/Alaska Native	1 (5.3%)	1 (1.3%)		0	0
Unknown	0	3 (3.8%)		1 (10%)	0
Relevant diagnoses
Obesity diagnosis	2 (10.5%)	6 (7.7%)	0.653	1 (10%)	4 (19.0%)	0.522
Snoring diagnosis	1 (5.3%)	1 (1.3%)	0.355	0	1 (4.8%)	0.483
Smoking diagnosis	0	0	0.355	0	1 (4.8%)	0.483
Alcohol diagnosis	0	0	0.355	0	1(4.8%)	0.483
Family History of HTN	17 (89.5%)	31 (39.7%)	0.0*	9 (90%)	9 (42.9%)	0.013*
Smoking status
Smoker or smoke exposure	5 (26.3%)	15 (19.2%)	0.532	7 (70%)	12 (57.1%)	0.492
Visit provider gender
Female	15 (78.9%)	47 (60.2%)	0.574	8 (80%)	15 (71.4%)	0.61
Attending provider gender
Female	12 (63.2%)	41 (52.5%)	1	6 (60%)	11 (52.4%)	0.69
Trainee involved (n = 57)
Yes	12 (63.2%)	49 (62.8%)	1	8 (80%)	11 (52.4%)	0.14
Trainee type (n = 57)
Resident
First year	5 (41.7%)	24 (49.0%)	0.123	5 (62.5%)	7 (63.1%)	0.138
Third year	1 (8.3%)	0		0	0
Fellow
First year	6 (50%)	25 (51.0%)	0.123	3 (37.5%)	1 (9.1%)	0.138
Second year	0	0		0	3 (27.3%)
Trainee to attending gender (n = 57)
Female, female	6 (50.0%)	29 (59.2%)	0.862	5 (62.5%)	5 (45.5%)	0.763
Female, male	5 (41.7%)	15 (30.6%)		2 (25.0%)	4 (36.4%)
Male, male	0	1 (2.0%)		1 (12.5%)	2 (18.2%)
Male, female	1 (8.3%)	4 (8.2%)		0	0
Trainee to attending gender match
Yes	6 (50.0%)	30 (31.2%)	0.526	6 (75.0%)	7 (63.6%)	0.599
Visit provider to patient gender match
Yes	11 (57.9%)	56 (71.8%)	0.274	4 (40.0%)	15 (71.4%)	0.093
BPA fired
Yes	4 (21.1%)	3 (3.8%)	0.026*	10 (100%)	13 (61.9%)	0.023*
Systolic BP HTN stage
2	3 (15.8%)	3 (3.8%)	0.019*	3 (30.0%)	1 (4.8%)	0.087
1	11 (57.9%)	67 (85.9%)		7 (70%)	17 (81.0%)
0	5 (26.3%)	8 (10.3%)		0	3 (14.3%)
Systolic any High BP
1	17 (89.5%)	71 (91.0%)	0.834	0	0
Diastolic BP HTN stage
2	1 (5.3%)	0	0.082	0	1 (4.8%)	0.746
1	5 (26.3%)	14 (17.9%)		2 (20%)	5 (23.8%)
0	13 (68.4%)	64 (82.1%)		8 (80.0%)	15 (71.4%)
Diastolic any high BP
1	8 (42.1%)	11 (57.9%)	0.036*	0	0
Any intervention done
Patient received an intervention	10 (52.6%)	30 (38.5%)	0.304	7 (70%)	15 (66.7%)	0.853

Diagnosis versus no diagnosis pre- and post-QI project (P < 0.05; ⱡp borderline significance). *P < 0.05; ⱡp borderline significance.

Younger patients (P < 0.001) and those with a family history of HTN (P = 0) were more likely to receive an HTN diagnosis. Only 3 other variables were significant in predicting HTN-related diagnoses: BPA firing (P < 0.026), the systolic BP stage (normal BP, stage 1 HTN, or stage 2 HTN) (P < 0.019), and the diastolic BP stage 1 (P < 0.036). Patient race trended toward significance (P = 0.058).

Age less than 18 years was the only variable significant in predicting whether an intervention (laboratory, referral, and imaging) occurred (P < 0.043) (Table 3).

Table 3.

Patients Meeting Criteria for Elevated BP or HTN Criteria

	Pre-QI Project			Post-QI Project
Patients with Elevated BP	n = 97			n = 31
Mean age	15.75 (P = 0.295)			15.24 (P = 0.049)*
BMI	28.13 (P = 1.47)			27.57 (0.267)
Any Intervention Done (Laboratory, Referral, Imaging)
	Intervention	No intervention	Sig	Intervention	No intervention	Sig
Patient gender
Female	14 (35.0%)	26 (65.0%)	0.993	7 (33.3%)	5 (50.0%)	0.373
Patient age (y)
Adolescent (<18)	34 (85.0%)	55 (96.5%)	0.043*	21 (100%)	7 (70.0%)	0.008*
Adult (18+)	6 (15.0%)	2 (3.5%)		0	3 (30.0%)
Ethnicity
Not Hispanic or Latino	22 (55.0%)	35 (61.4%)	0.211	12 (57.1%)	9 (90.0%)	0.067
Hispanic or Latino	18 (45.0%)	19 (33.3%)		9 (42.9%)	1 (10.0%)
Unknown/not reported	0	3 (5.3%)		0	0
Race
White	17 (42.5%)	19 (33.3%)	0.41	7 (33.3%)	4 (40.0%)	0.152
Other	13 (32.5%)	17 (29.8%)		8 (38.1%)	1 (10.0%)
Black/African American	9 (22.5%)	12 (21.1%)		3 (14.3%)	5 (50.0%)
More than one race	1 (2.5%)	4 (7.0%)		2 (9.5%)	0
American Indian/Alaska Native	0	2 (3.5%)		0	0
Unknown	0	3 (5.3%)		1 (4.8%)	0
Relevant diagnoses
Obesity diagnosis	4 (10.0%)	4 (7.0%)	0.714	1 (10.0%)	4 (19.0%)	0.522
Snoring diagnosis	0	2 (3.5%)	0.245	1 (4.8%)	0	0.483
Smoking diagnosis	0	0		1 (4.8%)	0	0.483
Alcohol diagnosis	0	0		1 (4.8%)	0	0.483
Family history of HTN	23 (57.5%)	25 (43.9%)	0.51	13 (61.9%)	5 (50.0%)	0.53
Smoking status
Smoker or smoke exposure	7 (17.5%)	13 (22.8%)	0.219	12 (57.1%)	1 (70.0%)	0.492
Visit provider gender
Female	11 (27.5%)	17 (29.8%)	0.825	15 (71.4%)	8 (80.0%)	0.61
Attending provider gender
Female	14 (35.0%)	24 (42.1%)	0.531	12 (57.1%)	5 (50.0%)	0.709
Trainee involved (n = 57)
Yes	28 (70.0%)	33 (57.9%)	0.287	13 (61.9%)	6 (60.0%)	0.919
Trainee type (n = 57)
Resident
First year	14 (50.0%)	15 (45.5%)	0.631	8 (61.5%)	4 (66.7%)	0.362
Third year	0	1 (3.0%)		0	0
Fellow
First year	14 (50.0%)	17 (51.5%)		2 (15.4%)	2 (33.3%)
Second year	0	0		3 (23.1%)	0
Trainee to attending gender (n = 57)
Female, female	17 (60.7%)	18 (54.5%)	0.451	7 (53.8%)	3 (50.0%)	0.311
Female, male	7 (25.0%)	13 (39.4%)		3 (23.1%)	3 (50.0%)
Male, male	1 (3.6%)	0		3 (23.1%)	0
Male, female	3 (10.7%)	2 (6.1%)		0	0
Trainee to attending gender match (n = 57)
Yes	18 (64.3%)	18 (54.5%)	0.602	10 (76.9%)	3 (50.0%)	0.241
Visit provider to patient gender match
Yes	25 (62.5%)	42 (73.7%)	0.27	12 (57.1%)	7 (70%)	0.492
BPA fired
Yes	2 (5.0%)	5 (8.8%)	0.696	16 (76.2%)	7 (70.0%)	0.713
Systolic BP HTN stage
2	3 (7.5%)	3 (5.3%)	0.882	4 (19.0%)	0	0.332
1	31 (77.5%)	47 (82.5%)		15 (71.4%)	9 (90%)
0	6 (15.0%)	7 (12.3%)		2 (9.5%)	1 (10%)
Systolic any High BP
1	36 (90.0%)	52 (91.2%)	0.837
Diastolic BP HTN stage
2	1 (2.5%)	0	0.482	0	1 (10%)	0.336
1	8 (20.0%)	11 (19.3%)		5 (23.8%)	2 (20%)
0	31 (77.5%)	46 (80.7%)		15 (76.2%)	7 (70%)
Diastolic any high BP
1	10 (25.0%)	13 (22.8%)	0.813	0	0
HTN/elevated BP diagnosis
Received accurate diagnosis	10 (25.0%)	9 (15.8%)	0.304	7 (33.3%)	3 (30.0%)	0.853

Intervention versus no intervention pre- and post-QI project. *P < 0.05.

BPA Intervention

The new BPA contained accurate BP guidelines and four separate BPA groups based on provider type, patient age, and HTN stage. One month after implementation, we ran a test analysis of the data to monitor appropriate firing. The BPA was not firing for resident-level providers; a month-delay postponed the postintervention chart review.

Investigators analyzed visits between April 13, 2017 and July 14, 2017 for BPA utilization, and there was a discrepancy between patients who had high BP and those whose BPA fired. Although the BPA should have fired for 8 patients, the BPA was rounding height percentiles up and down, resulting in 8 missed patients. All subsequent numbers were rounded down to capture all elevated BP and included the 8 missing patients in our data.

Postintervention Chart Review

A postintervention chart review occurred between April and July 2017. During this period, 6.1% (n = 31) of 461 patients had an elevated BP with the associated firing of the BPA (Table 1). Four patients (12.9%) met the criteria for stage 2 systolic HTN, 24 (77.4%) met the criteria for stage 1 systolic HTN, and 3 met the criteria for elevated Systolic BP. Seven patients (22.6%) met stage 1 diastolic HTN, and 1 patient (3.2%) met stage 2 diastolic HTN criteria.

The only factors significantly associated with the appropriate diagnosis were family history of HTN (P < 0.013) and firing of the BPA (P < 0.023). Age under 18 years was the only significant variable for predicting an intervention (P < 0.008) (Table 3).

DISCUSSION

Pediatric HTN frequently goes unnoticed in children and adolescents despite a growing prevalence. Our project evaluated possible causes for improper or lack of BP measurement and improved measurement, documentation, diagnosis, and treatment intervention within ambulatory care.

The study identified a lack of confidence in automated BP measurements. With a standardized protocol and educational reinforcements, confidence in manual and automatic BP improved while systolic and diastolic measurements decreased without the need to repeat measurements, fearing falsely elevated values. Reasons for this change are multifaceted, but some MA felt their training was not detailed enough during school. Many expressed little opportunity for practice as DinaMaps were standard in most clinical settings consistent with prior study findings.[12,13] With the extensive training and practice, the MAs went from “knowing” about manual BP measurement to “doing,” consistent with Miller’s Pyramid of assessment.[14]

There was little improvement with BPA usage, potentially due to fewer encounters secondary to improved BP measurement. It is also possible that the motivation for intervening with an elevated BP is different than recognizing the elevation itself. Providers could also be hesitant to intervene for an elevated BP due to knowledge or comfort with testing and BP medications. Prior studies with adult providers showed low adherence with evidence-based guidelines correlating to provider age, seniority, practice size, type of medical specialization, and increasing BP measurement.[15] Further research is warranted in the pediatric population.

The project’s limitations include focusing on one clinical setting in a specific period, restricting possible generalizability. We recognize that rotating trainees not familiar with clinic BP interventions, medications like stimulants for attention-deficit/hyperactivity disorder, and patients having laboratory testing for other conditions like prediabetes could affect outcomes.

As this project concluded, updated BP guidelines were released.[11] Despite changes for 13 through 17 year olds, we believe the project principles remain relevant. Education (continued over time for retention) and standardized training in BP measurements significantly impacted BP accuracy and reliability. Despite reliable vitals, providers frequently skip documentation, ignoring BPA reminders. Our clinic updated BPA alerts to align with the 2017 guidelines, and we continue to monitor BP documentation and interventions for elevated values.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENTS

Thank you to Rebecca Seale, MA, who acted as the study’s research assistant.