Lower peak expiratory flow rate is associated with a higher risk of pneumonia in patients with stroke

Authors: Wenxiu WU, Jingjing LIN, Xuezhen ZHOU, Suzhen YE, Mengmeng SHAO, Jiangying YU, Chengye ZHOU, Haiyan LI

Abstract

BACKGROUND

Low peak expiratory flow (PEF) rate is common in patients with stroke. Studies on changes in PEF rates in patients with stroke often have small sample sizes, limiting the generalizability of their findings.

AIM

This study aimed to compare the PEF rates between patients who were post-stroke with or without pneumonia and age- and sex-matched healthy controls and explore the PEF-pneumonia association among stroke survivors.

DESIGN

Prospective observational study.

SETTING

Department of Rehabilitation, the First Affiliated Hospital of Wenzhou Medical University.

POPULATION

Initially, 809 patients with stroke undergoing inpatient rehabilitation were recruited.

METHODS

Data collected included the demographics, stroke history, the presence of dysphagia, and the PEF rates on admission. Logistic regression analysis was conducted to identify the PEF threshold as predictive of pneumonia after adjusting for confounders.

RESULTS

Patients with stroke had a mean PEF rate of 243.89±139.38 L/min, significantly lower than that of the normal control group. The PEF rate was significantly lower in the pneumonia group than in the non-pneumonia group (P<0.001). Within the stroke cohort, the PEF rates were lower than the predicted rates (P<0.001). Older age, lower PEF(%),and dysphagia were associated with a higher pneumonia risk post-stroke per stepwise multivariate logistic regression analysis. Furthermore, the combination of these three significant predictors (PEF(%), swallowing function, and age) yielded an area under the curve of 0.857 .Regarding age, the cut-off point of ≥65.5 years was the optimal level to discriminate the presence of pneumonia among patients with stroke. For PEF%,the cut-off point of <60% was the optimal level to discriminate the presence of pneumonia among patients with stroke. For screening dysphagia, the patients with impaired safety only and those with impaired safety and efficacy faced a higher pneumonia risk.

CONCLUSIONS

Patients with stroke exhibited significantly lower peak expiratory flow rates compared to healthy controls after adjusting for age and sex and when compared to their reference values. Decreased PEF rates were independently associated with pneumonia development during inpatient rehabilitation in post-stroke patients.

CLINICAL REHABILITATION IMPACT

This study suggests that low PEF rates may predict pneumonia and that the prevention of PEF rate decline may prevent pneumonia development.

Stroke is an acute cerebrovascular disease that affects approximately 13.7 million individuals worldwide; It was the second most common cause of death and the third most common cause of disability in 2016.^1^ Previous studies have demonstrated that stroke affects the muscles of the upper and lower limbs and the respiratory system.^2^ Individuals with stroke typically exhibit alterations in breathing patterns,^3^ reduced ventilatory function,^4^ diminished strength of the respiratory muscles,^5^^, ^^6^ decreased activity of the paretic diaphragm,^7^^, ^^8^ and the onset of ineffective coughing and secretion retention.^9^^, ^^10^ Studies have shown that respiratory dysfunction after stroke leads to increased incidence of pneumonia, decreased exercise endurance, and increased risk of mortality.^11^^-^^13^ However, studies on changes in respiratory function in patients with stroke have often used small-sized populations, limiting the generalizability of their findings.

Most patients with stroke are bedridden in the acute phase due to paralysis, making bedside assessment the primary method for the prompt evaluation of respiratory function. The peak expiratory flow (PEF) rate is used to measure pulmonary function and is defined as the maximal flow generated with maximal force starting from total lung capacity.^14^ The PEF rate, to some extent, reflects the airway patency of an individual and is related to the individual’s effort level, lung capacity, and respiratory muscle strength.^15^^, ^^16^ It can be a valid option as a first-line screening tool for patients who cannot undergo spirometry. Recent research has indicated that the PEF rate can be a quantitative indicator of coughing ability.^17^ The measurement of the PEF rate is facilitated by a simple, affordable, user-friendly peak flow meter, which is particularly advantageous in environments with limited access to advanced equipment.^18^

This study aimed to examine the PEF rate of stroke survivors with or without pneumonia during inpatient rehabilitation, compared to age- and sex-matched healthy controls. Additionally, we aimed to investigate whether the PEF rate measured at admission could predict the risk of pneumonia development in stroke survivors. The findings may help inform the implementation of interventions aimed at improving respiratory function, thereby preventing further morbidity and mortality in patients with stroke.

Materials and methods

Clinical study design

This study used a prospective observational design. The study cohort comprised 809 consecutive patients who experienced a stroke and were admitted to the First Affiliated Hospital of Wenzhou Medical University for inpatient rehabilitation (January 2020-December 2022). All participants provided informed consent before enrollment. The study protocol was approved by the local ethics committee at the First Affiliated Hospital of Wenzhou Medical University (KY2022-093) and was conducted in accordance with the principles set forth in the Declaration of Helsinki.

Stroke diagnosis was confirmed through clinical and radiographic evaluations, with stroke defined as an acute cerebrovascular event resulting in focal or global dysfunction lasting >24 h. Following the completion of stroke treatment, all the patients discharged from the Department of Neurology were transferred to an inpatient rehabilitation unit. The inclusion criteria for enrollment in the study were age >18 years; confirmed diagnosis of stroke; hemiparesis; consciousness and cooperation; absence of facial paralysis hindering proper occlusion; absence of evident chest deformities; and ability to maintain a sitting position. The exclusion criteria included the absence of medical referral with a stroke diagnosis; age <18 years; the presence of previous comorbidities, such as epilepsy, significant cardiovascular disease, or malignant tumors; uncontrolled pneumonia at admission; and patients with concurrent clinical conditions such as dementia, severe cognitive impairment, receptive aphasia, loss of consciousness, visual or hearing impairments, severe facial paralysis, or poor cooperation interfering with evaluation. These exclusions were made to reduce the influence of individual variability on PEF. Additionally, control cases (N.=801) with an age and sex distribution matching that of the patients with stroke were included. The control patients had no diagnosed diseases and were not using any medication.

Data on baseline characteristics collected from the patients included demographic data (age, sex, height, and Body Mass Index [BMI]), history of risk factors (hypertension, diabetes mellitus, atrial fibrillation, hyperlipidemia, smoking habits, and alcohol abuse), and stroke-related factors (type and size of stroke and stroke severity according to the National Institute of Health Stroke Scale [NIHSS]). Additionally, the proportion of dysphagia, as assessed by the Modified Volume-Viscosity Swallow Test (V-VST),^19^ the proportion of unassisted standing, and the Barthel Index upon admission were documented. All patients received treatment according to current guidelines.

Peak expiratory flow

This study used a Mini Wright Peak Flow meter (Keka PEF-3) manufactured by Shanghai Marubo Science & Technology Co., Ltd. (Shanghai, China) following Chinese medical device standards. All materials used were disposable. The surfaces of commonly used equipment were decontaminated using 70% (w/v) rubbing alcohol. The PEF rate data were collected by two trained researchers to ensure that participants could execute the required actions correctly and reduce operational errors. All tests were conducted in the morning between 00 and 00.

The participants were given time to acclimate to the experimental and environmental conditions. Clear instructions about the test procedure were given to the participants, and a trial of the procedure was permitted to improve performance during the study. The participants were instructed to stand with the peak flow meter held horizontally in front of their mouths. They were then instructed to inhale deeply, firmly close their lips around the mouthpiece, and clamp their nose shut with a clip to prevent air leakage. Next, they were asked to exhale forcefully, and the number indicated by the cursor was recorded. This procedure was repeated twice, producing three readings at 2-minute intervals. The highest of the three values was considered the obtained PEF rate. The obtained PEF rate was compared with the reference rate established by Zhongnanshan et al. for the Chinese population^20^ using the following population-based

for predicted PEF rate (L/min) = 75.6 + 20.4 × A ‒ 0.41 × A2 + 0.002 × A3 + 1.19 × H (A: age; H: height);
for predicted PEF rate (L/min) = 282.0 + 1.79 × A ‒ 0.046 × A2 + 0.69 × H (A: age; H: height).

Modified volume-viscosity swallow test

The modified V-VST is based on the consumption of nectar, pudding, and liquid viscosities^19^ prepared by blending mineral water with a thickener following the manufacturer’s instructions. In this study, the patients were provided with the thickener Nutilis Clear^®^ (Nutrition Health Science, Switzerland). Nectar viscosity was first tested. Increasing volumes of 5, 10, and 15 mL were sequentially introduced into the patient’s mouth with a 50-mL syringe. If no alterations were detected in the safety parameters for nectar viscosity, the consumption of pudding and liquid viscosities were subsequently evaluated. Impaired labial sealing and oral residue led to impaired efficacy, while impaired safety was indicated by coughing and a decrease in oxygen saturation of ≥3%, measured using a finger pulse oximeter. The V-VST had a diagnostic sensitivity and specificity of 93.17% and 81.39% for dysphagia, respectively, and an inter-rater reliability (kappa) of 0.77.^21^

The patients were divided into four subgroups based on the comprehensive measurement of swallowing function, as 1) safety and efficacy, defined as patients presenting with no clinical signs of impaired safety or efficacy after completing 15 ml of three viscosity series; 2) impaired efficacy only, defined as patients presenting with clinical signs of impaired efficacy only after completing more than 5 mL of any viscosity series; 3) impaired safety only, defined as patients presenting with clinical signs of impaired safety only after completing more than 5 mL of any viscosity series; and 4) impaired safety and efficacy, defined as patients presenting with clinical signs of safety and efficacy after completing 5 mL of any viscosity series.

Impaired safety is more difficult for clinicians to deal with than impaired efficacy. If a patient presents with clinical signs of not only impaired safety but also impaired efficacy with different viscosity series, the results of the modified V-VST should be interpreted as clinical signs of impaired safety.

Diagnosis of post-stroke pneumonia

The diagnosis of clinically defined pneumonia was based on the Centers for Disease Control and Prevention criteria.^22^ These criteria necessitate the presence of a new and persistent infiltrate or consolidation on at least one chest radiography or computed tomography scan, along with one of the following clinical fever, leukopenia or leukocytosis, and altered mental status in individuals aged >70 years, in the absence of other causes. Additionally, these should occur with two of the following new-onset purulent sputum or changes in the characteristics of the sputum, new-onset or progressive cough, rales, and impaired gas exchange.

Statistical analysis

Data are presented as percentages or frequencies for discrete variables and medians (interquartile range) for continuous variables. Comparisons between groups were conducted using various statistical tests, including the Mann-Whitney U Test, one-way analysis of variance (ANOVA), Student’s t-test, Chi-square Test (χ^2^), and Fisher’s Exact Test, as appropriate. Univariate analysis was performed to assess the differences in the PEF rates when adjusted for age, sex, and height as covariates in the study. In cases where ANOVA indicated significant differences between groups, the post-hoc Tukey test was used for pairwise comparisons. Bonferroni corrections were applied to adjust for multiple comparisons. Stepwise multivariate logistic regression was performed, incorporating all factors that exhibited significant differences in the univariate analysis, such as age, sex, PEF, and mV-VST, to calculate the odds ratios (ORs) and corresponding 95% confidence intervals (CIs) for the occurrence of pneumonia following stroke. Receiver operating characteristic curve analysis was conducted to evaluate the accuracy of the pneumonia predictions and the optimum cut-off points. The level of statistical significance was set at a threshold of P <0.05. All statistical analyses were carried out using SPSS software (version 26; IBM Corp., Armonk, NY, USA).

Data availability

The original contributions presented in the study are included in the article; further inquiries can be directed to the corresponding author.

Results

Overall, 449 patients were excluded from the analysis based on the exclusion criteria. Data from 809 patients were analyzed (Figure 1). Table I presents the baseline characteristics of the 809 eligible patients, including the demographics, vascular risk factors, and clinical features. The patients’ mean (standard deviation) age was 63.19±11.90 years, and approximately 33.37% of the patients were women. Among them, 360 patients had dysphagia (44.51%) following the stroke event. Additionally, 35.11% (284/809) of the patients could not stand or required hands-on assistance at admission. During inpatient rehabilitation, 56 patients developed pneumonia, reflecting an incidence of 6.92% post-stroke. Compared to patients without pneumonia, those with pneumonia exhibited similar vascular risk factors and the most common baseline characteristics. However, the pneumonia group was older (72.05±10.35 vs. 62.53±11.75 years, P<0.001), had a higher NIHSS score (7.80±4.28 vs. 5.97±3.76, P=0.021), and had a lower Barthel index score (35.27±23.23 vs. 56.22±23.90, P<0.001). In patients with pneumonia, the proportion of those with atrial fibrillation was higher (17.86% vs. 5.44%, P<0.001), and the involvement of both cortical areas was more common (26.79% vs. 13.41%, P=0.006). Additionally, patients with pneumonia exhibited a higher rate of dysphagia (85.71% vs. 41.43%, P<0.001).

Figure 1: —Patient’s flow chart.

The mean PEF rate in patients with stroke was 243.89±139.38 L/min versus 371.52±114.78 L/min in healthy participants. Notably, the obtained PEF rate in patients with stroke was significantly lower than that of the healthy control group. The predicted PEF rate was higher than that of the healthy control group. Consequently, the obtained PEF level was significantly lower in patients with stroke, with or without pneumonia, compared to healthy controls (P<0.001). Furthermore, the obtained PEF levels were notably lower in the pneumonia group compared to the non-pneumonia group (P<0.001). Within the stroke groups, the obtained PEF rate was significantly lower than the predicted PEF rate according to the paired Student’s t-test (P<0.001). However, there was no significant difference between the obtained and predicted PEF rates in the normal group (P=0.128). Moreover, the PEF (%) of patients with stroke, with or without pneumonia, was significantly lower than that of the healthy control group. Additionally, the PEF (%) in the pneumonia group was notably lower than that of the non-pneumonia group (P<0.001). After adjusting for age, sex, and height, significant differences persisted between the obtained PEF, predicted PEF, and PEF (%) when comparing patients with stroke, with or without pneumonia, and healthy controls (all P <0.001) (Table I).

Table II presents the classification of the participants into three categories by PEF (%) at admission as <60%, from 60% to 80%, and >80%. PEF (%) <60% was the most prevalent category in the pneumonia and non-pneumonia groups of post-stroke patients (82.14% vs. 46.22%, respectively). Conversely, the proportion of patients with PEF (%) >80% was lower in the pneumonia group compared to the non-pneumonia group (8.93% vs. 29.74%, respectively). Consequently, the distribution of participants across the PEF (%) subgroups differed significantly between the pneumonia and non-pneumonia groups (P<0.001).

Older age, PEF (%), and swallowing function were identified as being associated with a higher risk of developing pneumonia post-stroke in the stepwise multivariate logistic regression analysis (Table III).

The risk of pneumonia was notably higher in patients with lower PEF or dysphagia compared to those without (OR=2.272, 95% CI: 1.386–3.724, P=0.001 and OR=2.237, 95% CI: 1.041-4.459, P<0.001, respectively). Furthermore, when the PEF (%) was below 60% in the logistic regression model with PEF (%) >80% as the reference, it was independently associated with pneumonia development (OR=3.580, 95% CI: 1.335 9.603, P=0.011). Regarding dysphagia, compared to the patients with normal swallowing function, both the patients with impaired safety only and those with impaired safety and efficacy faced a higher risk of developing pneumonia (OR=3.880, 95% CI: 1.639-9.188, P=0.002, and OR=11.741, 95% CI: 4.892-28.176, P<0.001, respectively). The combination of the three significant predictors (PEF (%), swallowing function, and age) yielded an area under the curve of 0.857 (95% CI: 0.803-0.911) (Figure 2).

Figure 2: —Receiver operating characteristic curve analysis with PEF (%), with PEF (%) + swallowing, and a model with the 3 PEF (%) + swallowing + age for post-stroke pneumonia. ROC: receiver operating characteristic; AUC: area under the ROC curve; CI: confidence interval; PEF: peak expiratory flow.

Regarding age, the cut-off point of ≥65.5 was the optimal level to discriminate the presence of pneumonia among patients with stroke. For PEF%, the cut-off point of <60% was the optimal level to discriminate the presence of pneumonia among patients with stroke. For screening dysphagia, the patients with impaired safety only and those with impaired safety and efficacy faced a higher risk of developing pneumonia.

Discussion

To the best of our knowledge, this study is the first to report a potential association between the PEF rate and the development of pneumonia in patients with stroke during their inpatient rehabilitation phase. This finding was consistent with previous large-scale studies that reported lower PEF rates as a risk factor for pneumonia.^23^^, ^^24^

In this study, the incidence of pneumonia during inpatient rehabilitation was 6.92%, which was lower than that in previous reports.^25^^, ^^26^ This discrepancy may stem from the fact that most prior studies were conducted in the neurology department, where post-stroke pneumonia was diagnosed within 7 days of stroke onset. In contrast, all the patients in our study underwent inpatient rehabilitation after completing acute stroke treatment and were followed up from admission to discharge, with 59.83% of the patients being admitted at least 10-30 days post-stroke. A comprehensive review of 54 trials reported varying incidences of stroke-associated pneumonia across different care settings, including 4.1-56.6% in neurological intensive care units, 17-50% in medical intensive care units, 3.9-44% in stroke units, 3.9-23.8% in the units of mixed studies, and 3.2-11% in rehabilitation units.^27^ Growing evidence suggests that specialized swallowing assessments and alternative feeding may mitigate the risk of stroke-associated pneumonia.^28^ Despite a substantial proportion of patients experiencing dysphagia following stroke in this study (44.51%), nearly all patients underwent swallowing assessments. They received alternative feeding advice from a speech and language pathologist upon admission. Therefore, this study excluded patients who presented with loss of consciousness, previous respiratory disease, or uncontrolled pneumonia at admission, as these factors are known to increase the risk of post-stroke pneumonia.^29^^, ^^30^ Frequency estimates in cohorts and trials exhibit wide variation, contingent upon patient characteristics and the timing and method of screening and treatment.

Neurological deficits from stroke induce muscle weakness throughout the affected hemibody, including the respiratory musculature.^2^ In 2013, a study^31^ involving 35 patients recovering from stroke and 35 controls for lung function testing showed a decreased forced lung capacity, forced expiratory volume in the first second, PEF rate, and chest expansion in hemiparetic individuals compared to controls (P<0.05). A study in 2017 also revealed that there was a reduction in respiratory muscle strength in the acute phase of stroke and that the respiratory muscle strength was lower in individuals with older age and higher BMI.^5^ The latest study indicates^6^ that the average maximum inspiratory and respiratory pressure in patients with stroke was significantly lower than the standard predicted values for the same age and sex, indicating respiratory muscle weakness in hemiparetic individuals due to stroke. Regarding forced spirometry, a significantly lower PEF rate was found in our study (243.89±139.38 L/min, 61.80±30.79% of the reference value) in participants who had a stroke compared to healthy participants (371.52±114.78 L/min, 98.14±22.47% of the reference value), after adjusting for age, sex, and height. Our results aligned with previous reports on decreased PEF rates in individuals who had experienced a stroke.^31^^, ^^32^

The PEF rates were significantly lower in patients with pneumonia than in those without pneumonia following a stroke. Our findings indicated that a low PEF rate predicted pneumonia development in patients with strokes during inpatient rehabilitation, aligning with the conclusions of prior studies. Respiratory disabilities post-stroke, linked with dysphagia and ineffective cough, may heighten the risk of pneumonia, recognized as the primary cause of non-vascular mortality post-stroke.^33^ PEF measurement reflects a mechanism shared with coughing, namely the rapid contraction of expiratory muscles resulting in decreased chest volume and swift exhalation of high-pressure gases.^17^^, ^^34^ Kulnik et al. demonstrated a correlation between an increase in the peak cough flow associated with the PEF^23^ and reduced pneumonia risk in patients with acute stroke who had swallowing dysfunction.^24^ Similarly, another study^35^ highlighted an association between low PEF and pneumonia development in individuals aged ≥65 years, many of whom had a history of stroke and neurology clinic visits. Jiang et al.^16^ found that objectively assessing cough strength through the PEF rate, a non-invasive and easily reproducible measure, predicts extubation failure in mechanically ventilated patients who have completed a spontaneous breathing trial. Additionally, some evidence suggests that expiratory muscle strength training may play a role in lowering the risk of respiratory infection.^11^^, ^^36^ Our study further revealed that the risk of post-stroke pneumonia varied based on the severity of the PEF. This underscores the utility of the PEF in preventing pneumonia among post-stroke patients, emphasizing its simplicity and repeatability as a predictive test for pneumonia development.

The relationship between dysphagia and pneumonia has been well recognized.^28^^, ^^37^^-^^39^ In a previous study, patients who failed dysphagia screening were found to be more likely to develop pneumonia (13.1% vs. 1.9%).^40^ Clave et al.^41^ specifically designed the V-VST to predict signs of dysphagia, such as safety and efficacy. A 2021 study^19^ revealed that the risk of aspiration pneumonia in patients with acute ischemic stroke varied according to the dysphagia severity. The pneumonia risk was higher in patients with dysphagia, regardless of whether safety and/or efficacy were impaired, compared to patients without dysphagia. Our research also showed that the risk of pneumonia was notably higher in patients with dysphagia compared to patients without (OR=2.237, 95% CI: 1.041-4.459, P<0.001). However, compared to patients with normal swallowing function, those with impaired safety only and those with impaired safety and efficacy faced a higher risk of developing pneumonia. The risks of pneumonia for patients with impaired efficacy only and those with normal swallowing function were similar. Therefore, further studies using more accurate evaluation methods, such as videofluoroscopy, are needed to confirm these findings.

Limitations of the study

This study has a few limitations. First, it was conducted in a single institution, leading to potential bias due to the single-center effect and the clustering of observations. Notably, the median NIHSS score was relatively low, suggesting a potential underestimation of the pneumonia incidence. This underestimation could be attributed, in part, to the exclusion of patients with severe stroke who had concurrent clinical conditions, such as dementia, severe cognitive impairment, receptive aphasia, or loss of consciousness, as per the study criteria. Sampling bias may also have existed, and the study could not reflect the entire stroke population. Second, the PEF test was performed with the participant seated, chosen for its practicality during measurement, compared to a standing position. It is worth acknowledging that PEF measurements are affected by various devices and measuring methods.^42^ The guidelines of international associations^43^ recommended conducting the maneuver in a standing position. However, there may be limitations and difficulties in performing standing measurements, especially for patients after a stroke.^44^ In contrast, some studies showed no significant differences between measurements taken in sitting or standing positions.^45^^-^^47^ A recent study^48^ revealed that when standing is impractical, conducting the measurements in a seated position may be reasonable. Thus, the study indicated a potential underestimation of approximately 9 L/min-1 or 2%, a difference deemed small and unlikely to substantially impact clinical management in most cases. Third, the modified V-VST was a screening test for dysphagia, in which it was hard to detect silent aspiration, potentially leading to an underestimation of the risk of pneumonia. Executing a videofluoroscopy or fiberoptic endoscopic evaluation of swallowing could be considered after a positive V-VST; however, it may not be feasible to perform an instrumental evaluation in every dysphagic patient. A recent review of the available evidence for using V-VST in the diagnosis of dysphagia reported good sensitivity and specificity, which may suffice as a diagnostic method, especially when instrumental tests are not readily available.^21^

Conclusions

In conclusion, the findings of this study demonstrate that, compared with healthy individuals, patients who have experienced a stroke exhibit significantly lower PEF rates, even after adjustment for age, sex, and height. Their PEF rates were also lower than the reference values. Reduced PEF is independently linked with the development of pneumonia in post-stroke patients during inpatient rehabilitation, indicating that low PEF can serve as a predictive marker for pneumonia. This finding can potentially help design preventive strategies aimed at PEF decline prevention to mitigate pneumonia risk. Finally, further prospective intervention studies on the effect of physical exercise or rehabilitation on improving PEF for pneumonia prevention are warranted.