Hospital‐based intervention to enhance hypertension diagnosis in Kalasin hospital, Thailand, 2017‐2019: A pre‐post pilot intervention study

Abstract

In Thailand 5.9 million individuals ≥15 years old have undiagnosed hypertension. The intervention to reduce undiagnosed hypertension was piloted and aimed to compare pre‐ and post‐intervention hypertension diagnosis rate and follow‐up rate. A hospital‐based pre‐ and post‐intervention study was piloted in a general hospital in Thailand. The intervention included an electronic pop‐up alert when raised blood pressure was observed and a follow‐up protocol. The follow‐up protocol entered patient information in a follow‐up book that scheduled an appointment to recheck blood pressure. Statistical analyses compared the rate of hypertension diagnosis and follow‐up between the pre‐ and post‐intervention periods, adjusted for differences in baseline characteristics. A post‐intervention, self‐report survey among nurses and nurse‐aids explored perceptions about raised blood pressure management and solicited suggestions to improve the intervention. 574 raised blood pressure patients visited the hospital in the pre‐intervention period; 27.4% returned for follow‐‐up; and hypertension diagnosis rate was 1.4%. Among 686 post‐intervention raised blood pressure patients, overall hypertension diagnosis rate improved to 6.1%. In per‐protocol patients, 81.9% were booked to follow‐‐up, hypertension diagnosis rate was 18.6%, and the adjusted odds ratio of hypertension diagnosis was 4.5 times higher compared with the pre‐intervention period. By self‐report, 20% of health workers had no time to provide the follow‐up book due to work overload, yet >57% reported that information technology improved detection of raised blood pressure and improved follow‐up. The interventions significantly increased the hypertension diagnosis rate and follow‐up among raised blood pressure patients in a single hospital but may benefit from incorporating an information technology‐assisted follow‐up protocol.

1. INTRODUCTION

Hypertension (HTN) is one of the leading risk factors for global mortality, ¹ estimated to have caused 10.6 million deaths and 143 million DALYs lost in 2015. ² HTN is a major cardiovascular risk factor, and uncontrolled HTN may lead to stroke, myocardial infarction, or cardiac failure. ³ , ⁴ Globally, the World Health Organization set a goal of a 25% relative reduction in the prevalence of HTN. In Thailand, HTN prevalence among adults aged ≥15 years was 21.4% in 2009; the national goal is to lower prevalence below 16% by 2025. Despite this goal, HTN prevalence rose in the short term from 2009 to 2014 to 24.7%. ⁵

Of the estimated 13.3 million hypertensive Thai adults, about 5.9 million are undiagnosed. ⁶ Community screening could identify only 0.8 million of these patients. ⁷ As a consequence, hospital‐based, opportunistic screening may identify more undiagnosed and untreated HTN cases, which in turn could increase HTN treatment and control, ultimately leading to public health gains. A preliminary investigation found that> 6.4 million government hospitals visitors were likely to have RBP. About 2.7 million of them had persistent high BP but without any HTN diagnosis recorded. ⁸

Most of studies about screening HTN in Thailand have studied community‐based screening; few have studied the feasibility of opportunistic, hospital‐based screening. This pilot study aimed to develop an intervention to improve HTN diagnostic rate by increasing hospital‐based screening efforts and linkage to HTN treatment using a pre‐ and post‐intervention design.

2. METHODS

2.1. The intervention

Figure 1 provides a summary overview of patient flow. The intervention was comprised of two components: an automated electronic pop‐up message and a follow‐up protocol. For the pop‐up message, the BP of a patient was recorded (these BP recordings could be entered by hospital health care workers as part of routine examination for any presentation at the hospital in initial steps before patients meet physicians) in the electronic medical record of the hospital information system. When the BP met criteria of RBP (SBP ≥ 140 mm Hg and/or DBP ≥ 90 mm Hg), a message to the health care provider instantly appeared recommending a follow‐up evaluation. The nurse gave an appointment card to RBP patients instructing them to return for a separate appointment to recheck their BP.

Figure 1.

Guideline for follow‐up RBP patients in this study

The follow‐up book included one page of appointment detail, with three chemical copy papers, ensuring that the original follow‐up record was copied onto the two remaining pages and retained at the hospital. Nurses recorded patient name and surname, phone number, address, BP, and date and place of follow‐up appointment. Each time when health care workers (HCWs) entered information, the two copy pages were kept on file to be a source of information for tracking the follow‐up process. The original pages were shared with the patients. The backside of the follow‐up book listed suggestions and inform patients about HTN and its potential health consequences and the importance of follow‐up.

The follow‐up sites were all 24 primary health care centers under the catchment area of the hospital. The hospital was the biggest in the province with around 560 beds. The outpatient department (OPD) served around 1216 cases/day. Complicated patients from every district of the province were referred to the hospital. Coming to the hospital and waiting for a long time only to recheck BP must be a factor that discourages follow‐up. Therefore, for follow‐up, less crowded health care centers, near the patient’s home or worksite, were selected.

Eligible patients were entered into the follow‐up book and scheduled for an appointment to recheck BP in following two weeks. If patients still had RBP at the first follow‐up appointment but were not yet diagnosed, they would be scheduled for a second follow‐up appointment to recheck BP within two weeks. If patients failed to come on the appointment date, we would phone them to remind them to follow‐up. If those patients could promise follow‐up within seven days, they would return to their local clinic. But if they could not give the certain date within seven days or they missed their appointment village health volunteers (VHVs) would visit their house and arrange follow‐up. The follow‐up could occur within 1‐6 weeks.

2.2. Inclusion and exclusion criteria

We included patients aged ≥15 years who lived in the catchment area of the hospital and had SBP ≥140 mm Hg and/or DBP ≥ 90 mm Hg on at least one BP measurement in the hospital (defined as RBP), identified in the electronic medical record of the hospital. Furthermore, eligible patients were those without a history of HTN, defined as having no chart record of HTN diagnosed by physician, no visit coding to an ICD‐10 code related to HTN (I10 – I16), and no record of prescription for antihypertensive drugs from history taking nor in HIS. We excluded patients with BP <140/90 mm Hg, as well as those who had visited the emergency room and pregnant women.

2.3. Study design and study site

The design was hospital‐based pre‐intervention and post‐intervention study. This study was implemented in a provincial hospital in Kalasin province, in northeastern part of Thailand. In the background, there were no major changes in HTN screening and management practice guideline by the provincial health office nor the hospital during 2017 – 2019. One exception was that in 2019, the provincial health office extended their outpatient service to the nearest health care centers so as to transfer non‐complicated patients out of the hospital. Normal practice of BP measurement in the hospital was for patients to self‐measure BP in a kiosk in the presence of a hospital observer. If BP was higher than 140/90 mm Hg on the kiosk machine, nurses in each department remeasured BP by a digital automatic device or a manual sphygmomanometer after a five‐minute rest. RBP recorded in the electronic medical record of the hospital information system represented the lowest BP among those measured directly by the nurse observer upon the repeat measurement (after at least five‐minutes rest and at least one remeasurement by the nurse, if the nurse had time for more than one BP measurement, she entered the last BP she observed into the electronic medical record).

2.4. Study period

The study observation interval was divided into a pre‐intervention period during December 15, 2017 – March 15, 2018 (last follow‐up date, April 26, 2018), when the hospital did their routine service without any intervention about HTN screening, and a post‐intervention period during December 15, 2018 – March 15, 2019 (the last follow‐up date, April 26, 2019), when we implemented the intervention.

2.5. Operational definitions

The 2019 Thai Guidelines on The Treatment of Hypertension ⁹ defines patients into four groups: high normal BP (mean BP ≥ 130/80 mm Hg.), possible HTN (mean BP ≥ 140/90 mm Hg.), probable HTN (mean BP ≥ 160/100 mm Hg.), and definite HTN (mean BP ≥ 180/110 mm Hg). Patients with high normal BP, possible HTN or probable HTN, and any of target organ damage, cardiovascular diseases, diabetes mellitus, or high cardiovascular risk are diagnosed immediately. Patients without any of the above conditions may be referred for home BP monitoring, ambulatory BP monitoring, or serial office BP monitoring in order to make a more definitive diagnosis. The Thai HTN treatment protocol has two components: 1) lifestyle modification and 2) antihypertensive drugs. Blood pressure lowering treatment is extended to patients meeting one of two criteria: average office BP 130‐139/85‐89 mm Hg along with known cardiovascular disease, or BP ≥ 140/90 mm Hg regardless of cardiovascular disease status. Patients without known cardiovascular diseases, renal disease, or target organ damage may be prescribed a trial of lifestyle modification for 3‐6 months, before medication treatment is considered. Patients with BP ≥ 160‐179/100‐109 mm Hg are eligible to start antihypertensive drugs as soon as they get diagnosed.

Based on Thai hypertension guidelines, RBP patients were classified into 2 RBP stages. Stage I was SBP = 140‐159 mm Hg. and DBP = 90‐99 mm Hg. Stage II was SBP ≥ 160 mm Hg. or DBP ≥ 100 mm Hg.

In follow‐up process, we defined four categories of patients who did not come to follow‐up as either “missed appointment date” (any patient who did not come to follow‐up places at the date of appointment), “receded” (any patient who did not appear in the hospital Health Information System (HIS) again after they had RBP recorded within study period), or “lost follow‐up” (any patient who cannot be contacted by phone called nor home visited), or “declined” (any patient who rejected to continue follow‐up program).

First follow‐up rate was defined as number of patients who came back to follow‐up in the first follow‐up visit divided by number of total eligible patients referred by the follow‐up book.

2.6. The intervention implementation

Hospital information technology staff developed a pop‐up message alert and pilot tested it before starting the post‐intervention period. We had a meeting with focal point persons in the hospital to orient and train them on the intervention and also planned the method by which nurses would enter appointments into the follow‐up book. Focal point persons had to in turn orient other HCWs to the intervention in their department. We also had a meeting with focal point persons of each of the follow‐up health care centers to orient them to the intervention and instructed them on with the way to follow‐up patients.

2.7. Data collection

In the pre‐intervention period, we retrieved data from the HIS of the hospital and HIS of all follow‐up health care centers by retrieving data from the provincial health office database.

In post‐intervention period, FPP in the hospital collected the follow‐up book duplicate papers from each department and keyed‐in the data into an electronic spreadsheet. This spreadsheet was then sent electronically to the Principal Investigator of the project weekly. The Principal Investigator checked follow‐up status by matching the follow‐up date from the duplicate copy papers and actual follow‐up data from health care centers. Then, study staff called patients who missed appointments and sent a list of patients who missed appointment back to each health center. HTN diagnosis data in all follow‐up health care centers also retrieved data from the provincial health office database.

Retrieved variables included unique citizen identification, sex, weight, height, BP, occupation, visit date, visited department, and diagnosis code.

2.8. Ethics consideration

This study protocol was approved by Kalasin Hospital Research Ethics Committee (KLSH REC). Only aggregated and de‐identifed data were analyzed.

2.9. Statistical analysis

We summarized included RBP patients’ characteristics as means or percentages and tested for differences between pre and post‐intervention characteristics, using a t test for continuous variables and chi‐square test for categorical variables. For P‐values, all tests were two‐tailed. Multivariable logistic regression estimated adjusted odds ratio (95% CI) of HTN diagnosed at follow‐up after adjusting for age, sex, BMI status, RBP status at the first detected visit, and occupation. Intention‐to‐treat analysis estimated the adjusted odds ratio of HTN diagnosed in the post‐intervention period compared with the pre‐intervention period overall (whether or not protocol was followed). Standard errors are obtained using the delta method.

2.10. Post‐implemented intervention survey

Self‐reported questionnaires were distributed to hospital HCWs who delivered the study intervention, to explore perception about the risks of RBP and suggestions to improve the intervention.

3. RESULTS

3.1. Characteristics of the study population in pre‐intervention and post‐intervention periods

In pre‐intervention period, over 15 000 patients with at least one RBP measurement were entered into the HIS. After applying our exclusion criteria, 589 RBP patients remained; of these, 15 were diagnosed with HTN at the same day of the RBP measurement. This left 574 possible HTN patients eligible for follow‐up evaluation for diagnosis. More than half of the pre‐intervention period RBP patients had visited either the internal medicine or general outpatient departments; the remainder had visited other departments (surgery, dental, counseling, or orthopedics).

In post‐intervention period, again >15 000 patients had at least one RBP measurement; only 686 RBP patients remained for study after exclusion criteria were applied. Of these 686 included patients, only 188 had their names entered into the follow‐up book (“per protocol”). The general pattern of distribution of visits among generalist and specialist clinics remained similar to the pre‐intervention period.

Age, sex, and BMI status were not different between patients in pre‐intervention and post‐intervention periods (Table 1). Among post‐intervention period patients who were not entered into the follow‐up book (33.3%), the number of patients who came to the first follow‐up visit was only slightly higher than patients observed in pre‐intervention period (27.4%). However, for patients in post‐intervention period who were entered into the follow‐up book per protocol, first follow‐up visit rate was dramatically higher (81.9%).

Table 1.

Characteristics of eligible patients between pre‐intervention and post‐intervention period

Characteristics	Pre‐intervention (N = 574)		Post‐intervention (N = 686)
	No.	%	Patients not entered into follow‐up book (N = 498)			Patients entered into follow‐up book (N = 188)
			No.	%	P value a (compared with pre‐ intervention)	No.	%	P value b (compared with pre‐intervention)	P value c (compared with post‐intervention, no follow‐up book)
Age (mean)	53 y		53 y		.81	53 y		.70	.83
Age group
15‐44	144	25.1	121	24.30	.95	41	21.8	.40	.52
45‐59	231	40.2	204	40.96		86	45.7
≥60 y	199	34.7	173	34.74		61	32.5
Sex
Male	255	44.4	233	46.8	.44	96	51.1	.11	.32
BP status at first visit
Stage II	95	16.6	77	15.5	.63	49	26.1	<.01	<.01
Follow‐up visit, until know the result (HTN or not)
Yes	143	24.9	167	33.5	<.01	138	73.4	<.01	<.01

BMI status	(N = 573)		(N = 493)
Overweight and obese	338	59.0	312	63.3	.15	117	62.2	.43	.80
Occupation
Unemployed	103	17.9	109	21.9	.17	25	13.3	<.01	<.01
Blue collar	389	67.8	311	62.5		151	80.3
White and gray collar	82	14.3	78	15.7		12	6.4
Diabetes
Yes	47	8.2	40	8.0	.93	4	2.1	<.01	<.01

Abbreviation: CI, confidence interval; No., number.

^a P‐value for significant difference for pre‐ and post‐intervention comparing post‐intervention subgroup patients who did not get follow‐up book to total pre‐intervention patients based on t test for continuous variable and chi‐square for categorical variable; all tests two‐tailed.

^b P‐value for significant difference for pre‐ and post‐intervention comparing post‐intervention subgroup patients who got follow‐up book to total pre‐intervention patients based on t test for continuous variable and chi‐square for categorical variable; all tests two‐tailed

^c P‐value for significant difference for two subgroups at post‐intervention based on t test for continuous variable and chi‐square for categorical variable: all tests two‐tailed.

3.2. Hypertension diagnosis in pre‐intervention and post‐intervention periods

In pre‐intervention period, of the total of 574 eligible patients, 16.6% (95/574) of them initially had RBP stage II; however, only 157 of 574 followed up and 83.4% (131/157) of this group was found to be normotensive on repeat blood pressure examination. The follow‐up rate was 27.4% (157/574) and overall HTN diagnostic rate was 1.4% (8/574) (Figure 2.)

Figure 2.

Results of follow‐up in the pre‐intervention period

In the post‐intervention period, among 498 eligible patients who were not entered into the follow‐up book, 15.5% (77/498) of them had RBP stage II, but again only 166 of 498 followed up and 83.2% (138/166) were normotensive on further blood pressure testing. Among 28 patients who were advised to return to the hospital for a repeated screening blood pressure measurement, 13 patients returned to the hospital for a blood pressure recheck. Of these, seven were diagnosed with hypertension, resulting in a diagnostic rate of 1.4% (7/498).

Among 188 eligible post‐intervention patients who were entered into the follow‐up book per protocol, 34 patients were lost to follow‐up, 58.4% (90/154) were normotensive, and 13 were diagnosed with HTN upon second blood pressure measurement. Among the remaining 51 patients who were advised to return to the hospital for a third screening blood pressure measurement, 38 came to the hospital and 22 were diagnosed as hypertensive. The overall hypertension diagnosis rate was 18.6% (35/188) in post‐intervention patients entered into the follow‐up book (Figure 3).

Figure 3.

Results of follow‐up in the post‐intervention period

Overall, inclusive of per follow‐up book protocol and off‐protocol RBP patients, hypertension diagnosis rate in the post‐intervention period was 6.1% (42/686).

3.3. Effect of the follow‐up book intervention on multivariable‐adjusted probability of hypertension diagnosis

After adjustment for age, sex, BMI status, BP status at the first RBP observation, and occupation, the adjusted odds of hypertension diagnosis among patients in post‐intervention period who did not receive the follow‐up book was not significantly different from the pre‐intervention period comparator [adjusted odds ratio (AOR) = 1.10 (95% CI 0.39‐3.10)]. The adjusted odds of hypertension diagnosis in post‐intervention patients who received the follow‐up book was significantly higher compared with the pre‐intervention period [AOR = 16.47 (95% CI 7.33‐37.02)]. The overall AOR among all patients in post‐intervention period (with or without follow‐up book; “intention to treat”) was 4.54 compared with the pre‐intervention period (95% CI 2.13 – 9.66) (Table 2.)

Table 2.

Probability of hypertension diagnosis in the post‐intervention period compared with the pre‐intervention period

Exposure factors	Adjusted odds ratio among patients without follow‐up book (95% CI)	Adjusted odds ratio among patients with follow‐up book (95% CI)	Adjusted odds ratio among all patients in post‐intervention period (95% CI)
Overall intervention effect	1.10 (0.41‐3.01)	16.47 (7.07‐38.38)	4.54 (2.13‐9.66)
Sex
Male	0.65 (0.15‐2.76)	14.86 (5.44‐40.6)	4.11 (1.55‐10.90)
Female	1.61 (0.36‐7.28)	17.33 (4.82‐62.30)	5.27 (1.53‐18.14)
Age group
15‐44	1.19 (0.17‐8.60)	3.64 (0.50‐26.68)	1.80 (0.32‐9.96)
45‐59	1.13 (0.70‐18.23)	60.88 (7.98‐464.38)	16.13 (2.14‐121.38)
≥60 y	0.92 (0.24‐3.48)	12.65 (4.37‐36.60)	3.42 (1.26‐9.36)
BMI group
Non‐overweight	0.65 (0.12‐3.56)	11.45 (3.66‐37.74)	3.40 (1.10‐10.47)
Overweight and obese	1.36 (0.36‐5.11)	20.43 (6.90‐60.53)	5.83 (2.02‐16.80)
BP at the first visit
Stage I	1.14 (0.36‐3.56)	19.00 (7.66‐47.13)	4.93 (2.05‐11.89)
Stage II	0.61 (0.05‐6.88)	9.07 (1.85‐44.60)	3.58 (0.75‐16.96)
Occupation
Not working	–	–	–
Blue collar	0.78 (0.25‐2.40)	9.45 (4.16‐21.48)	3.31 (1.50‐7.30)
White or gray collar	–	–	–

Estimates are derived from multivariate logistic regression analysis adjusted for age, sex, BMI status, BP status at the first RBP observation, and occupation. For all comparisons, the reference group is the pre‐intervention phase cohort. Not working, and White or gray collar subgroup of occupation could not calculate due to omitted.

3.4. Post‐intervention survey results

The response rate among nurses involved in the intervention was 83% (40/48). Firstly, we asked their opinions about the appropriateness and experience of making follow‐up appointments with RBP patients. Of the nurses, 57% responded that all RBP patients should be scheduled for an appointment for recheck BP. Three quarters (31/40) of nurses responded they had seen the follow‐up book while the rest stated they had never seen it before. Among the nurses who reported seeing the follow‐up book, 10 of them never entered a patient into it. Furthermore, 20% of all nurses reported they could not provide the follow‐up book to all eligible patients because it competed with time demands of their main job and the nurse did not know how to make time to enter patients into the follow‐up book. Twenty‐one nurses estimated that the average time to make a follow‐up book appointment was around 5 minutes/case. More than half responded that automated detection of RBP by the hospital information technology assisted in detecting RBP patients, but that another automated system for making follow‐up appointments, replacing the paper follow‐up book, would be more helpful for diagnosing hypertension due to limited time to make an appointment.

4. DISCUSSION

We found that when a follow‐up book intervention was coupled with automated detection of RBP patients, the HTN diagnosis rate increased from 1.4% to 18.6%. In previous studies, among 126 patients who were identified as RBP in emergency department and were invited for follow up in 30 days, 51 patients had follow up and 39 patients had HTN, the HTN diagnosis rate was 30.95% (39/126). ¹⁰ In another study, the HTN prevalence among patients with RBP identified in the emergency department was around 31.6% when follow‐up was performed with home blood pressure monitoring. ¹¹ A third study revealed a prevalence of 24.4% on follow‐up of patients identified with RBP in a dental department. ¹² Compared with the studies that recruited from the emergency department studies, our HTN prevalence was lower, but also closer to the prevalence from the dental department study that was more similar to the hospital‐based ambulatory care environment reflected in our study.

The adjusted odds ratio of HTN diagnosis in the post‐intervention period among patients who received the follow‐up book was 16 times compared with patients in pre‐intervention period. However, this rate may not represent the true diagnostic yield of the intervention, as 27.4% (188/686) of post‐intervention patients were not entered into the follow‐up book. The “intention‐to‐treat” adjusted odds ratio when all patients in post‐intervention period are included was 4.5. Further study is needed to explore the system, patient, or provider factors that explain why some patients were not entered into the follow‐up book. It is possible that nurse orientation to this intervention was inadequate; the post‐intervention survey found that almost a quarter (22.5%) of involved health care workers had never seen the follow‐up book before and among those who saw the follow‐up book, 25.0% of them had never used it. Nurses’ limited time and crowded hospital clinics could explain this last finding. Despite the fact that many patients were not entered into the follow‐up book, the >4.7% percentage point overall increase in HTN diagnosis compares favorably with a past HTN screening quality improvement study from the United States. ¹³

Based on the results observed in this study, a hospital‐based HTN screening intervention could increase the HTN diagnostic rate substantially in Thailand. There are no financial or logistical barriers for accessing hospital care in Thailand; thus, a hospital‐based opportunistic screening program would likely achieve good population coverage over time. Second, once RBP patients are referred for definitive hypertension diagnosis and care, primary care in Thailand is quite strong. Every sub‐district has a community health center staffed with nurse practitioners who can initiate treatment for uncomplicated hypertension. Furthermore, Thailand has over a million village health volunteers (VHVs) trained in basic health knowledge and practices. ¹⁴ Both nurse practitioners and VHVs are playing an important role in the hypertension follow‐up process.

RBP measured at a single visit may not represent the patient’s true clinic blood pressure, ¹⁵ , ¹⁶ which can only be revealed with repeated measures of blood pressure over time (“regression to the mean” phenomenon). ¹⁷ , ¹⁸ , ¹⁹ In our study, the rate of normotensive blood pressure following an initial RBP was much higher than that reported in previous studies. ¹¹ , ²⁰ Transient environmental, behavioral, and measurement technique factors may lead to artificially high blood pressure on first measurement. The intervention hospital is a general hospital, serving more than 1000 patients daily. This patient volume may lead to inadequate blood pressure measurement technique. For example, a previous study showed that simply providing adequate rest time (at least five minutes) led to awake ambulatory blood pressures (mean 139.4/80.7 mm Hg) or automated office blood pressures (mean 140.5/83.1 mm Hg) significantly lower than a single office BP measured by nurse (mean 155.1/90.2 mm Hg). ²¹

The proportion of patients who had RBP stage II among patients in post‐intervention period who received the follow‐up book was 26.1%, which was substantially higher than the follow‐up book entry rate for RBP stage I (15.5%). Hypertension unawareness among HCWs could be a reason ²² ; 17.5% of those nurses reported no need to follow up if blood pressure is near to 140/90 mm Hg. In a previous study, health care workers reported that if patients had no symptoms, they would perform BP re‐assessment only if blood pressure was greater 162/95 mm Hg (nurses) or 169/100 mm Hg (physicians). ¹⁵ This unofficial practice could why RBP stage II triggered almost twice the rate of follow‐up book rate compared with RBP stage I.

5. LIMITATIONS

There were several limitations in this study. First, we could not get more accurate HTN diagnostic rate among all RBP patients due to low follow‐up visit rate, especially among the patients not referred for the follow‐up book. Second, health worker deployment of the intervention was not universal, because we could not force all HCWs to agree and fully participate in this study. Third, blood pressure measurement in crowded hospital outpatient clinics possibly caused a high rate of false positive of RBP at the time of first assessment. Fourth, as a single‐center pilot study, the results of this study may not be generalizable to other hospitals in Thailand or in other countries.

6. RECOMMENDATIONS

If the results of this single‐center study can be replicated, the implication will be that hospital‐based, opportunistic HTN screening should be implemented to efficiently increase rate of making HTN diagnosis and subsequent treatment and control. However, before testing this intervention more broadly, the intervention design should be modified. Making the follow‐up book process simpler, perhaps by using an IT‐assisted, automated protocol might increase ease of use for, and acceptance by, HCWs. Even if the technology improves, it will still be important to ensure appropriate education and protocol orientation to health care workers in both the screening and follow‐up sites before implementing the intervention more broadly. In 2020, Division of Non‐Communicable Diseases, Department of Disease Control, Thailand, will pilot adopt a version of the RBP identification and follow‐up intervention in four provinces in Thailand. If this approach continues to yield similar efficacy when implemented at other hospitals, this would justify scale up to the national level. Implementing this intervention in the national level will expand result of searching population at risk of HTN in Thailand from not only from community screening but also hospital‐based screening, more HTN screening coverage. Further research is needed to develop a user‐friendly information technology tool to support patient registration and tracking and allow secure sharing data among health care workers.

7. CONCLUSION

Opportunistic screening may improve HTN diagnosis in Thailand. However currently, there is no system for follow‐up of patients presenting with RBP in outpatient hospital clinics, which represents a missed opportunity to improve HTN diagnosis and prevent cardiovascular disease deaths. A hospital‐based intervention consisting of identifying RBP followed by systematic follow‐up of patients with RBP can significantly increase the rate of HTN diagnosis.

CONFLICT OF INTEREST

None.

AUTHOR CONTRIBUTIONS

Khanuengnij Yueayai and Phanthanee Thitichai conceived and designed the study. Khanuengnij Yueayai, Piyanut Pratipanwat, Siwaboon Chaisongkram, and Ladda Anosri designed and implemented the intervention process. Khanuengnij Yueaya collected, analysed, and interpreted data and prepared draft of the manuscript. Andrew E. Moran critically revised the manuscript. Phanthanee Thitichai supervised the study.

Yueayai K, Moran AE, Pratipanwat P, Chaisongkram S, Anosri L, Thitichai P. Hospital‐based intervention to enhance hypertension diagnosis in hospital K, Thailand, 2017‐2019: A pre‐post pilot intervention study. J Clin Hypertens. 2020;22:1310–1320. 10.1111/jch.13953