Biomedical engineering

Information for the general public

Biomedical EngineeringDespite current supervision as many as 25% of patients receiving haemodialysis treatment for endÿstage kidney failure experience deleterious side effects including loss of blood pressure, dizziness, cramps or sickness during a treatment. During dialysis treatment we currently monitor a patient's temperature, pulse rate, blood pressure, fluid content and weight, and sometimes also add information about their breathing rate and blood oxygen content. However, this data is collected separately and analysed separately. We are studying in collaboration with colleagues in Oxford's Institute of Biomedical Engineering whether by integrating these data (data fusion) we can more accurately predict which patients will have problems during dialysis.

Information for students

We are currently running a pilot study in conjunction with the Institute for Biomedical Engineering in Oxford.

Study Title

Can integrated vital sign monitoring contribute to the care of haemodialysis patients?

Study Design

Observational pilot study

Study Participants

Patients receiving regular maintenance hospital haemodialysis

Planned Sample Size

40 patients

Follow-up duration

Each patient will be monitored on 8 dialyses over a 6 month period - ideally they will be monitored on 4 consecutive dialyses on 2 occasions separated by about 3 months.

Planned Study Period

Data will be collected over a 6 month period. Initial data analysis will occur over the next 6 months, but data will be retained for re-analysis over the following 5 years.

Primary Objective

To assess whether integrated vital sign monitoring improve the sensitivity and specificity of predicting patient deterioration during a haemodialysis session.

Secondary Objectives

  1. Whether an alternative measurement, the pulse transit time, either on its own, or in combination with estimates of the patients blood volume status, can be used as a surrogate marker for blood pressure during dialysis.
  2. Whether changes in the peripheral pulse wave form (measured electronically) are indicative of changes in blood supply to the patient's periphery.
  3. We want to assess correlation between the data fusion scores in a patient and that patient's subjective views of their dialysis treatment.
  4. By studying the same patient repeatedly and a cohort of individual different patients we wish to identify intra-patient and inter-patient differences over time.
  5. We wish to determine whether it is possible to use the data fusion model to predict intra-dialytic morbid events such as drop in blood pressure, and subsequently modify dialysis to tailor this treatment for a particular patient in particular circumstances.
  6. We will also assess the tolerability and the acceptability of extra monitoring being included in this pilot study to patients receiving haemodialysis.
  7. We wish to assess whether the Visensia index helps to distinguish between patients already clinically judged to be stable and unstable.
  8. We wish to see whether training the data fusion interpretation on multiple exposures of a single patient to haemodialysis will allow it to become more sensitive and specific in picking up deterioration in that individual, than using a population based training algorithm.

Intervention (s)

Patients will be monitored non-invasively before, during and after routine haemodialysis sessions. They will also be asked to complete a questionnaire to assess their symptomatology.