Purpose
Potential harm to patients and departmental delays were identified in MRI brain studies involving cerebrospinal fluid shunts. This prompted a review and change in departmental protocol.
Externally programmable cerebrospinal fluid shunts allow for non-invasive management of cerebrospinal fluid pressure via assessment and adjustment of an external valve (Fig 1). Up to 80% of valves are affected by MRI fields, potentially resulting in uncontrolled changes in intracranial pressure.[1]
Pre-MRI documentation of shunt type and setting can reduce patient harm and improve departmental efficiency.[2] At the Royal...
Methods and materials
All MRI brain studies involving cerebrospinal fluid shunts from September 2017 – April 2018 were filtered from the Royal Brisbane and Women’s Hospital imaging database. Clinician request forms and patient-completed safety questionnaires were retrospectively analysed for patient demographics, shunt type, pre- and post-MRI shunt settings, and accuracy of documentation.
A clinical standard was set: 95% of MRI request forms for cerebrospinal fluid shunt patients should state shunt type and setting; less than 80% of shunt settings should require correction by radiography staff after MRI.
Results
There were 42 MRI studiesinvolving cerebrospinal fluid shunts (Fig 2). 33.3% of request forms were incorrectly completed, stating that no programmable shunt was present when patients did in fact have one. 59.5% of request forms did not document shunt type or setting. There were 5 recorded instances of shunt settings being altered by MRI exposure (Table 1).
Conclusion
Theclinical standard for MRI request forms was not met. A large proportion were incorrectly completed or lacked adequate information.
Patient safety was compromised. During a 7-month period at our centre, ≥5 shunt settings were altered by MRI. This number may even be under-represented by our data as post-MRI settings were not recorded for every patient. Undetected changes to shunt settings can result in over or underdrainage of cerebrospinal fluid with potentially catastrophic consequences.
Interestingly, 97.5% of patients correctly self-identifiedthe presence of a shunt compared to...
Personal information
Denver Khoo, Resident Medical Officer, Royal Brisbane and Women’s Hospital
Supun Abeyratne, Resident Medical Officer, Redcliffe Hospital
Kate Mahady, Neurointerventional Radiologist, Royal Brisbane and Women’s Hospital
References
1. Zabramski JM, Preul MC, Debbins J, McCusker DJ. 3T magnetic resonance imaging testing of externally programmable shunt valves. Surgical neurology international. 2012;3:81.
2. Giancardo L, Arevalo O, Tenreiro A, Riascos R, Bonfante E. MRI Compatibility: Automatic Brain Shunt Valve Recognition using Feature Engineering and Deep Convolutional Neural Networks. Scientific reports. 2018;8(1):16052.
3. Miyake H. Shunt Devices for the Treatment of Adult Hydrocephalus: Recent Progress and Characteristics. Neurol Med Chir (Tokyo). 2016;56(5):274-83.