Improving adherence to pulmonary hypertension screening in patients with systemic sclerosis: Overcoming the provider-level barriers

Abstract

Background:

Pulmonary arterial hypertension is a life-limiting complication in patients with systemic sclerosis. Current recommendations suggest a minimum screening of an annual transthoracic echocardiogram and pulmonary function test. We hypothesize that modifiable provider-level hurdles contribute to inadequate screening and can be alleviated.

Methods and Results:

We performed a longitudinal study over 23 months of all systemic sclerosis patients seen in pulmonary and/or rheumatology clinics. Pulmonary arterial hypertension screening was measured monthly as percentage of patients with up-to-date transthoracic echocardiogram and pulmonary function test out of the total systemic sclerosis patients with an outpatient encounter that month. Tests were considered “up-to-date” if they were performed 12 months prior or within the next 6 months post-encounter. Baseline adherence to PAH screening was assessed using a 9-month preproject period. Physicians in rheumatology and pulmonary clinics were surveyed for knowledge and perceived as barriers to screening. Interventions focused on provider-level hurdles to improve the screening rate to >90%. Adherence to systemic sclerosis–associated pulmonary arterial hypertension was compared between baseline and post-intervention phases as a continuous variable using nonparametric Mann–Whitney–Wilcoxon test and as categorical variable using chi-square test for overall adherence rate. A total of 18 providers were surveyed, of which 67% knew minimal screening recommendations and 44% identified “difficulty ordering tests” as a barrier. Our interventions were as follows: (1) didactic education and literature dissemination and (2) best practice alert in the electronic medical record to facilitate just-in-time ordering of screening. Adherence was measured for 23 months, comprising 166 systemic sclerosis patients. Monthly adherence to pulmonary arterial hypertension screening increased by 10.4% (69.2%–79.6%, p = 0.01 with chi-square and p = 0.09 with Mann–Whitney–Wilcoxon test), with an increase in transthoracic echocardiogram by 7.5% and pulmonary function test by 4.9%.

Conclusion:

Provider-level hurdles contribute to sub-optimal systemic sclerosis–associated pulmonary arterial hypertension screening. Lack of knowledge and difficulty in ordering tests were major barriers. Structured education and decision-support aids improve screening. These interventions, however, were insufficient to improve screening to a goal target of >90%.

Introduction

Pulmonary arterial hypertension (PAH) is a life-limiting complication of systemic sclerosis (SSc) that affects an estimated 1 in 12 patients with SSc. ¹ Despite the overall improvement in PAH survival, systemic sclerosis–associated pulmonary arterial hypertension (SSc-PAH) carries a poor prognosis. ² Early detection and subsequent treatment of SSc-PAH lead to improved survival. ³ Early symptoms of SSc-PAH are nonspecific, thus symptoms-based screening delays diagnosis. ⁴

Screening recommendations for SSc-PAH vary ranging from an annual transthoracic echocardiogram (TTE) to a multitude of tests that include TTE, pulmonary functions tests (PFTs), and laboratory testing.^{5 –8} Screening with an annual TTE alone is insufficient for PAH detection.^9,10 The use of PFT with diffusing capacity of lung for carbon monoxide (DLCO) is a good predictor for PAH in SSc. ¹¹ Most screening recommendations incorporate both TTE and PFT for annual SSc-PAH screening.^{5 –8}

Despite the benefits of early detection of SSc-PAH, appropriate screening is often impeded by provider-level barriers and patient factors. Mitigating provider-level barriers through systemic interventions, such as just-in-time reminders, and ease of execution have improved screening for other conditions. ¹² In this prospective, quality improvement study, we tested our hypothesis that SSc-PAH screening can be improved by overcoming provider-level barriers.

Methods

Setting and study population

University of Cincinnati Medical Center (UCMC) is an academic health system located in Cincinnati, Ohio. It is a nationally designated pulmonary hypertension association care center and scleroderma center. All of its outpatient rheumatology and pulmonary clinics were included in the study.

All patients who had an outpatient encounter in rheumatology or pulmonary clinics at UCMC were eligible. Patients with SSc, identified using International Classification of Diseases–10th Revision (ICD-10) codes (M34.9), without a pre-existing PAH diagnosis, were included as potential subjects.

Study design

Defining and measuring PAH screening

We reviewed published literature for screening recommendations that are summarized in Table 1.^{5 –8} Given the differences in screening recommendations, consensus by local experts, including rheumatology and pulmonary physicians, deemed an annual TTE and PFT that include DLCO and appropriate SSc-PAH screening.

Table 1.

Comparison of PAH screening guidelines in patients with SSc.

SSc-PAH screening guidelines/ recommendation source	Specific tests	Frequency
European Society of Cardiology/European Respiratory Society guidelines for the diagnosis and treatment of PAH 5	TTE, PFT with DLCO, and biomarkers (nonspecific: NTproBNP and BNP suggested)	Annual
DETECT study for detections of PAH in SSc 6	PFT, physical exam, EKG, and biomarkers (NTproBNP, anti-centromere antibody, serum urate) Followed by TTE if initial screening is positive	NA
American College of Rheumatology consensus recommendations for screening and detection of connective tissue disease–associated PAH 7	Initial PFT with DLCO, TTE, and serum NTproBNP Followed by annual PFT with DLCO and TTE Repeat PFT with DLCO, TTE, and serum NTproBNP if new symptoms develop or worsen	Annual
Australian Scleroderma Interest Group (ASIG) algorithm for PAH screening in patients with SSc 8	PFT and serum NTproBNP Follow-up TTE if initial screening is positive	6–12 months

TTE: transthoracic echocardiogram; PFT: pulmonary function test; DLCO: diffusing capacity of lung for carbon monoxide; EKG: electrocardiography; PAH: pulmonary arterial hypertension; SSc: systemic sclerosis; NTproBNP: N-terminal (NT)-pro hormone B-type natriuretic peptide; NA: not applicable.

PAH screening was measured monthly as the percentage of patients with up-to-date TTE and PFT out of the total SSc patients without PAH who had an outpatient rheumatology or pulmonology encounter during that month. Tests were considered “up-to-date” if they were performed within the previous 12 months or subsequent 6 months after the encounter. The post-encounter period was included to account for pragmatic delays in obtaining TTE or PFT after it was ordered in the clinic. Additional data included site of encounter and patient demographics.

Data were collected retrospectively at two separate points. Once before any intervention to determine baseline adherence to PAH screening using a 9-month preproject period (February 2016 to October 2016). Second, at 14 months after initial interventions to determine adherence throughout the intervention phase (November 2016 to June 2017) and the post-intervention phase (July 2017 to December 2017). Monthly adherence rates for all clinic encounters were calculated to assess the affect of interventions over time. Adherence was also assessed for each patient (patient-level) during each phase of study.

Identifying provider-level barriers to SSc-PAH screening

We performed a cross-sectional survey of rheumatology and pulmonary providers at UCMC to assess their knowledge regarding screening guidelines, testing intervals, and benefits of screening. They were also asked about their practice and perceived barriers to screening. The 16-question survey was distributed to providers in print and online format (online supplement).

Interventions

Study interventions were designed based upon identified provider-level barriers with an aim to increase adherence to appropriate SSc-PAH screening. The major barriers were lack of knowledge, variation in practices, and difficulty in ordering tests. These factors were addressed by two key interventions: (1) provider education through didactic session and dissemination of educational materials and (2) development of an electronic medical record (EMR)-based best practice alert (BPA) for just-in-time reminder and facilitation in ordering appropriate screening tests.

For provider education, a national expert was brought to UCMC through the Pulmonary Hypertension Association for a lecture to pulmonary and rheumatology faculty. The session focused on the various screening guidelines and benefits of early SSc-PAH detection. Next, an educational tool including materials covering guidelines and benefits of SSc-PAH screening was disseminated among all pulmonary and rheumatology providers.

We developed an EMR-based BPA to notify pulmonary and rheumatology providers of a patient’s need for SSc-PAH screening. The BPA was designed to display during in-person, outpatient rheumatology and pulmonary clinic encounters for patients with a charted diagnosis of SSc who do not have a TTE and/or PFT within 12 months of the encounter date. The BPA also provided links to the appropriate orders to facilitate screening (online supplement).

Statistical analysis

We performed descriptive analysis for patient characteristics in our population and for the results of the provider survey. The primary outcome of monthly adherence to appropriate SSc-PAH screening and its individual components (TEE alone, PFT alone, and TTE + PFT combined) was compared between baseline period and post-intervention phase as a continuous variable using nonparametric Mann–Whitney–Wilcoxon (MWW) test and as categorical variable using chi-square test for overall adherence rate. A p-value less than 0.05 was considered significant.

Results

Cross-sectional provider survey

A total of 18 providers participated (8 rheumatologists and 10 pulmonologists), of which 72% knew that SSc-PAH screening improves patient outcomes while the remaining 28% thought the data were inconclusive. Furthermore, 50% knew that early diagnosis of SSc-PAH improves mortality; 67% of providers knew the minimal screening recommendations for SSc-PAH; 28% would recommend screening only for symptomatic or high-risk SSc patients, while 72% recommend screening all patients; and 44% identified “difficulty ordering the correct screening test(s)” as a barrier to screening (online supplement).

Patient characteristics/demographics and encounter information

A total of 166 discrete patients were seen in the 23-month data collection period. Their characteristics are available in online supplement.

In the 23-month time period, there were a total of 837 outpatient clinic encounters with SSc patients with 379 (45.3%) in pulmonary clinics and 458 (54.7%) in rheumatology clinics.

SSc-PAH screening pre- and post-intervention

During the baseline period, the average (±standard deviation) monthly adherence to appropriate SSc-PAH screening (combined TTE and PFT) was 69.2 (±8.3)%, while TTE alone was 79.5 (±7.3)% and PFT alone was 81.2 (±8.8)%. When comparing the baseline period to post-intervention phase, the monthly adherence to appropriate SSc-PAH screening increased by 10.4% (from 69.2% to 79.6%, p = 0.006 with chi-square and p = 0.05 with MWW), while the increase in TTE alone was 7.5% (from 79.5% to 87%, p = 0.01, p = 0.09) and PFT alone was 4.9% (from 81.2% to 86.1%, p = 0.10, p = 0.31) (Figure 1).

Figure 1.

Adherence to pulmonary arterial hypertension (PAH) screening in patients with systemic sclerosis (SSc). Adherence to appropriate screening is defined as having up-to-date TTE and PFT tests within the 12-month pre-encounter to 6-month post-encounter time frame.

TTE: transthoracic echocardiogram, PFT: pulmonary function test, MWW: Mann–Whitney–Wilcoxon test.

*p-value less than 0.05 was considered significant.

When data were analyzed by individual patients, both the rates of TTE (75% vs 83%) and PFT (70% vs 79%) were higher in post-intervention group compared to baseline group, however not statistically different. The combination of TTE + PFT was significantly higher in post-intervention phase compared to baseline (58% vs 72%, p = 0.021) (Table 2).

Table 2.

Adherence rates to PAH screening tests by individual patient.

	Baseline period	Post-intervention period	p-value*
No. of individual patients, n	129	112
TTE, n (%)	97 (75.2)	93 (83)	0.137
PFT, n (%)	91 (70.5)	89 (79.4)	0.112
TTE + PFT, n (%)	75 (58.1)	81 (72.3)	0.021

TTE: transthoracic echocardiogram, PFT: pulmonary function test.

Adherence was assessed based on last clinical encounter by each patient during the respective study phase. The last months of both study phases were 14 months apart to avoid cumulative rate effect in post-intervention phase.

^*Chi-square test with p-value less than 0.05 was considered significant.

Discussion

Early detection and treatment of PAH among patients with SSc improve mortality.^2,3 Provider-level barriers contribute to sub-optimal SSc-PAH screening, leading to preventable patient harm. Our findings indicate that providers lack knowledge regarding value, timing, and components of SSc-PAH screening. Furthermore, providers have difficulty ordering appropriate tests. Our interventions were designed to mitigate provider-level hurdles to screening and comprised concentrated education and a just-in-time EMR prompt that facilitated the ordering of appropriate tests. These interventions improved the rate of SSc-PAH screening by >10% with sustainable results beyond the initial intervention phase without adding any further cost or resources. While we observed improved care, our interventions were insufficient to reach the target adherence of >90%, suggesting a role of other system factors beyond the scope of our project.

To the best of our knowledge, this is the first study aimed at improving SSc-PAH screening among SSc patients. While previous studies have shown that screening programs lead to earlier PAH diagnosis and improved survival, ⁴ no study has tested methods to improve adherence in real-world settings. Prior studies exploring SSc-PAH screening have reported similar screening adherence rates to our baseline as well as similar, significant discrepancies in adherence to screening among providers.^13,14 These findings highlight the wide variations in providers’ practices. Our study further corroborates inadequate SSc-PAH screening, even in specialized tertiary care centers, and proposes simple, yet effective, interventions to improve care. It is conceivable that such interventions will have a greater impact on health systems where baseline adherence is much lower.

We found the most common provider-level barriers to screening to be deficient in knowledge regarding screening guidelines and its benefits, as well as difficulty in ordering appropriate tests. These barriers were similar to those previously described. ¹⁴ A lack of consensus for SSc-PAH screening likely contributes to this gap in knowledge. Similar provider-level barriers have been noted in lung cancer screening, specifically gaps in providers’ knowledge regarding the specific guidelines and benefits of screening.^15,16 Structured education focusing on the evidence behind screening and its appropriate application and benefits lead to provider “buy in” and improved screening. ¹⁷ While such education-based intervention is important, knowledge decay is a known occurrence in the absence of repeated exposure. ¹⁸ In our study, we provided education through expert-led didactics followed by dissemination of an educational tool for repeated review. Notably, participation was not mandatory or measured afterwards, thus its effect on the long-term sustainability is unclear.

Education or dissemination of guidelines alone is not sufficient for sustainable impact. ¹⁹ The best results in lung cancer screening were achieved by an approach that incorporated EMR-based clinical reminders and real-time feedback on provider’s screening performance in addition to education. ¹⁶ Our intervention included structured education followed by implementation of EMR-based BPA designed to be context specific, to provide just-in-time reminders, and to facilitate ordering appropriate screening tests. BPAs have been effective in improving screening for other disease conditions, such as hepatitis C screening, with improvements in screening rates from 7.6% to 72%. ¹² We also achieved improved adherence rates reaching 79% (for combined TTE and PFT) that sustained beyond the initial phase, suggesting a possible limit of such EMR-based, provider-centered interventions.

There are several possibilities to why a further increase to >90% was not achieved with our current intervention. While BPAs are created with best intentions to assist providers and improve care, the EMR is crowded with alerts that may cause “alert fatigue.” ²⁰ To allow for variation in care driven by patient factors (not provider factors), our BPA was a “soft reminder” allowing it to be overturned or ignored. This function may have reduced effectiveness if the BPA was canceled without reviewing its content. Developing a method to track providers’ BPA performance and provide feedback would be helpful and could potentially improve adherence to BPA prompts; however, due to resource limitations, we were unable to create such a feedback loop.

Our study had several limitations. First, this is a single center at a specialized, tertiary care center in the United States. However, our provider-level barriers are not unique and similar interventions will likely improve SSc-PAH screening adherence for other centers as well. Second, the patients were identified using retrospective EMR data using ICD codes, which has known accuracy limitations. Importantly, this inherent bias was stable throughout the study period and would not affect the improvements seen in outcomes. Since our focus was to improve adherence at the level of health system, we did not obtain detailed case-to-case data to take into account appropriate denial/deferral to screening such as prognosis or patient preference. Finally, the scope of our project was limited to improving SSc-PAH screening adherence, and we did not address the factors that contribute to delay in confirmatory diagnosis or initiation of therapy once patients are screened positive.

Conclusion

Early detection of PAH through annual screening in SSc patients improves mortality; however, provider-level barriers hinder appropriate screening. One major barrier is the knowledge deficit regarding the method and benefits of screening. Lack of consensus among guidelines likely contributes to this knowledge gap. Structured education coupled with “just-in-time” decision-support aids led to sustained, improved rates of SSc-PAH screening. While these interventions effectively address provider-level barriers, further efforts are needed to address patient and healthcare delivery factors. Adherence to SSc-PAH screening should be considered as a quality performance indicator for centers caring for SSc patients.