Evaluating the Feasibility of a Piezoelectric Smart Sensor as an Aid to Help Clinicians Screen for the Development of Pneumonia in an At-risk Population

This is a proof of concept study to demonstrate that the piezoelectric sensor can be used as an aid to help clinicians screen for the development of pneumonia in at-risk populations (i.e. hospitalized patients with acute stroke and COVID-19 patients). The findings from this study can be extrapolated to the other community settings (e.g. nursing homes, dormitories) where access to healthcare is more limited. The decision to site this study in an acute hospital or community care facility (CTF) is largely pragmatic - easier access to patients; investigations (e.g. chest x-ray, blood tests) and treatment are more readily available for patients without any need for their transfer.

As this is a proof of concept study, the project number of subject required was not based on statistical power calculation. A total of 70 subjects will be recruited. The sample size of 30 participants is a pragmatic decision based on feasibility and is typical for a proof of concept study of this nature.

A total of 70 subjects, including those with acute stroke (with MRI or CT brain evidence of acute stroke) at NUH, those with ESRF (End Stage Renal Failure) on HD (hemodialysis) at NUH, The piezoelectric sensor will be applied to the posterior chest wall (6-point) of the subject at one time point daily (excluding weekends, public holidays and discharge day) up to 2 weeks whilst he/she is warded. CXR (Chest X-Ray) will be performed if the clinician and/or piezoelectric sensor detects abnormal chest sounds that is suggestive of pneumonia or fluid overload.

The piezoelectric sensor will be applied to the skin of the subject at the 6 points on the posterior chest wall for 30 seconds at each position. The piezoelectric sensors will be carefully sterilized with alcohol swaps before and after each placement. The program algorithm will then determine the type of lung sound recorded by the piezoelectric sensors.

CXR will be performed to determine the presence of pneumonia or fluid overload if the clinician and/or piezoelectric smart sensor detects abnormal lung sounds. CXR may not be performed if it is not required based on primary team's clinical assessment.

The piezoelectric sensor will be applied to the posterior chest wall of the subject at one time point daily (excluding weekends, public holidays and discharge day) up to 2 weeks whilst he/she is warded, until pneumonia or fluid overload is detected and confirmed by CXR or the subject is discharged, whichever is earlier. In the event that the subject declines to have the piezoelectric sensor placed every day, a minimum of 50% of compliance rate is required.

If the piezoelectric sensor detects abnormal lung sounds while CXR is normal, the sensor will continue to be placed on the subject daily until CXR confirms pneumonia or fluid overload or the subject is discharged, whichever is earlier.

A standard posterior to anterior view CXR will be obtained if the piezoelectric sensor detects abnormal lung sounds and the CXR will be interpreted using a standardized approach.

Inclusion criteria

• Aged 21 years and above
• Acute stroke patients (i.e. MRI or CT brain findings consistent with diagnosis) or ESRF patients on HD who are mentally competent and cognitively fit to provide consent.

Exclusion criteria

• Individuals who are unable to give informed consent
• Individuals with active dermatological conditions