Results
Seven hundred thirty-three patients with ICH received neurosurgical consultation through the Telbios telemedicine system at the neurosurgical clinic of the University of Messina between June 2003 and June 2011. These patients with ICH represented 26% of all patients who needed an urgent neurosurgical consultation through telemedicine during the selected period (2819 patients).
The first end point of this study was the analysis of the time frame needed to provide a neurosurgical second opinion consultation. Analyzing the CT scanning records and the relevant neurosurgical opinion, we recorded a mean of 38 minutes (range 23–109 minutes) between hospital admission at the spoke and neurosurgical consultation at the hub. The mean time was 160 minutes (range 68–204 minutes) for a second opinion consultation obtained through patient transportation, before the introduction of the telemedicine system. The time needed to obtain neurosurgical teleconsultation at the periphery was not dissimilar to that in Messina's urban area, which was estimated at an average of 26 minutes (range 0–37 minutes).
The second study end point was the analysis of primary and secondary transfers to our neurosurgical department and neurointensive care. Data are summarized in Fig. 1. During the evaluation period, 176 (24%) of 733 patients needed a primary transfer to the academic hub for the management of ICH. Selecting patients for transfer was based on a combination of clinical, demographic, and radiological characteristics. Hematoma location, CT features, and size were the radiological characteristics considered in combination with the age of, relevant clinical data on, and neurological status of the patient. In cases of controversial interpretation, the neuroradiological team at the hub reevaluated the images.
(Enlarge Image)
Figure 1.
Algorithm summarizing referral strategies for patient selection. SAH = subarachnoid hemorrhage.
On the basis of hematoma location, capsular, thalamic, or posterior fossa ICH in patients older than 45 years and with a history of systemic hypertension was presumptively judged as primary hypertensive ICH, without the need for further diagnostic workup. Transfer to the hub of patients with ICHs in the aforementioned locations was done on the basis of hematoma size and patient clinical status (Fig. 1). Patients with capsular or thalamic ICHs with volumes ranging between 30 and 70 ml—that is, with evidence of relevant mass effect—and a GCS score > 8 were transferred to the hub and underwent surgical treatment (Fig. 2). With regard to posterior fossa ICHs, compression of the fourth ventricle and/or a clot size > 3 cm were considered criteria for transfer to our neurosurgical unit. Patients underwent surgical treatment if they had a GCS score < 13 or evidence of neurological deterioration.
(Enlarge Image)
Figure 2.
Telemedicine screenshot showing multiple slices of brain CTs obtained in a 45-year-old patient with spontaneous nucleocapsular ICH. Telemedicine allowed rapid visualization of the CT and the collection of relevant clinical data. In general, nucleocapsular hematoma was not managed by central transfer, unless there was evidence of relevant mass effect with possible deterioration requiring neurointensive care or surgical control of ICP. Cautious conservative management led to a good outcome in this case.
Patients with blood in the entire ventricular system were considered at risk for hydrocephalus and were centralized, whereas patients with only a minimal amount of blood in 1 lateral ventricle were left at the peripheral hospital and observed. All patients with a subarachnoid clot in the basal or sylvian cisterns associated with temporal or IVH were admitted to the central hospital for further studies. Patients with brainstem hemorrhage were admitted only if the GCS score was > 8 to treat or prevent deterioration as a result of acute hydrocephalus. Patients in a deep coma because of large brainstem injuries were treated only with supportive care.
With regard to lobar hemorrhages, patients younger than 50 years of age were admitted to the central hospital (Fig. 3). Older patients with evidence of vasculopathy remained at the peripheral hospital with the indication for performing MRI and MRA.
(Enlarge Image)
Figure 3.
Upper: Brain CT scans obtained in a 27-year-old patient with spontaneous lobar ICH. Telemedicine screenshot reproduces the CT console of the peripheral hospital with visualization of the scouts, multiple slices of the examination, and tools for image management. In general, hematoma without evidence of mass effect was managed at the peripheral hospital, unless one suspected vascular malformations, tumors, or sinus thrombosis, requiring urgent management. In such cases, clinical data collection is fundamental to avoid disastrous delay of diagnosis and treatment. Lower: Angiograms showing evidence of a dural arteriovenous fistula (left), which was promptly managed with embolization (right).
In general, other CT features were considered, including marks suggestive of vascular abnormalities, tumors, sinus thrombosis, or hemorrhagic transformation of ischemic stroke, and were evaluated on an individual basis. When a secondary ICH was suspected, the patient was admitted to the neurosurgical department to complete our institutional diagnostic workup protocol consisting of CTA, MRI, MRA, SPECT, and eventually angiography.
The 557 patients who remained at a peripheral hospital had neuroradiological images and clinical/neurological conditions that did not require direct neurosurgical or neurointensive care management. They were treated through clinical, neurological, and neuroradiological monitoring, according to the neurosurgical consultation and indications. In particular, the neurosurgeon recommended a control CT scan 6 hours later and a second control scan 24 hours after stroke onset. This practice was different for patients directly admitted to the central hospital, who did not undergo the 6-hour CT control study unless their condition clinically deteriorated.
Eight (1.4%) of 557 patients needed a secondary transfer to our department for a worsening of clinical conditions and/or deterioration of neuroradiological findings. Five of these 8 patients were on anticoagulation drugs for other diseases. Patients in a deep coma or with a moribund status were not transferred, however, and were given only supportive care.
Of the 176 patients who were centralized, 95 (13% of the overall group) underwent surgical evacuation of an ICH. Surgical indication was individualized and based on the patient's age and neurological status, hematoma size and location, and related mass effect (Fig. 4). The remaining 81 patients (11%) underwent clinical, neurological, and neuroradiological monitoring and received only medical therapy. Thirty-nine of these patients underwent further studies for suspected vascular malformations, tumors, or sinus thrombosis. Suspicion was confirmed in 21 patients. In terms of efficacy of interpretation, clinical and radiological data were misinterpreted in 26 cases, including 8 cases of secondary centralization and 18 cases of inappropriate centralization, corresponding to 3.5% of the whole series.
(Enlarge Image)
Figure 4.
Upper: Computed tomography scans obtained with the telemedicine system, showing ICH and IVH in a 24-year-old patient. Cases of large lobar hemorrhage in young patients were centralized to undergo emergency CTA and surgical treatment. Lower: Computed tomography scans obtained after emergency surgical treatment.