Thrombolysis in Postoperative Stroke

Nicolas Voelkel, MD; Nikolai Dominik Hubert, MSc; Roland Backhaus, MD; Roman Ludwig Haberl, MD; Gordian Jan Hubert, MD


Stroke. 2017;48(11):3034-3039. 

In This Article


Data Collection

The Telemedical Project for Integrative Stroke Care (TEMPiS) is a TeleStroke network with 2 neurological stroke centers (hubs) and 19 regional hospitals (spokes) in Southeast Bavaria/Germany. Currently 9000 patients/y receive treatment within the network. Detailed descriptions of the concept were published previously.[7,8]

Local stroke care is provided by internal medicine departments (nonstroke specialists) in 13 hospitals and by neurological departments in 6 hospitals. All hospitals have access to stroke expertise via telemedicine 24/7. Hospitals with neurological departments use the teleconsultation service mainly to obtain a second opinion.

The TEMPiS Teleconsultation Registry is a prospective registry of all teleconsultations performed within the TEMPiS network including clinical data and recommendation on IVT.

The TEMPiS IVT registry includes all consecutive patients receiving IVT. All network hospitals use a uniform IVT protocol for documentation of times (onset, admission, and treatment), National Institute of Health Stroke Scale (NIHSS) and formal contraindications. All protocols are sent to the network coordination center for plausibility checks before entering the data into the registry. Data from February 2003 to October 2014 were searched for IVT treatment and keywords for any preceding surgery. Patients were included if surgery had been performed within 90 days before stroke. Discharge letters of all identified patients were collected from the network hospitals and reviewed.

Data security was approved by German authorities. As part of quality assurance measures, the registries did not need approval of the institutional review board.

Definitions for Subgroup Analysis

Recently and Nonrecently Performed Operations. Recently performed operations were defined as surgical interventions 1 to 10 days before IVT. When performed between day 11 and 90, surgery was defined as nonrecently.

Major and Minor Surgery. Exact definition of the terms major and minor in classifying surgery remains controversial. To reflect surgical bleeding risk, we adapted the classification as suggested by Pilcher[9] by including surgery in well-vascularized tissue (ie, colon mucosa, and prostate tissue) and surgery of large arteries in the major group.

Major Surgery. Major surgery involves opening major body cavities—abdomen (laparotomy), chest (thoracotomy), or skull (craniotomy)—or surgery in well-vascularized tissues or of any large arteries. Vital organs can be stressed. The surgery usually is done by a team of doctors using general anesthesia in a hospital operating room. A stay of at least 1 night in the hospital usually is needed after major surgery.[9]

Minor Surgery. In minor surgery, major body cavities are not opened, well-vascularized tissue or large arteries are not affected. Minor surgery can involve the use of local, regional, or general anesthesia and may be done in an emergency department, an ambulatory surgical center, or a doctor's office. Vital organs usually are not stressed, and surgery can be done by a single doctor, who may or may not be a surgeon. Usually, the person can return home on the same day that minor surgery is done.[9]

All surgical procedures of our cohort were independently classified as major surgery and minor surgery by 2 of the coauthors (NV and GJH). In cases of conflicting initial categorization, agreement was sought through discussion.

If a patient had received >1 surgery within 90 days before IVT, the patient was categorized for statistical analysis to the more recent/more serious surgery group. For example, the patient was categorized as presenting with a recent surgery if at least 1 surgery had been performed within the 10-day time window. Accordingly, the subject was categorized as presenting with a major surgery if at least 1 surgery had been major.

Outcome Measures

Primary outcome measure was SSH.

Secondary outcome measures were intracranial hemorrhage (ICH), bleeding complications in other areas, and in-hospital mortality.

Definitions for Outcome Parameters

Surgical Site Hemorrhage. SSH was defined as any bleeding in the surgical wound region, along the same compartment or a nearby structure.

Bleeding Severity. Classification of bleeding severity was performed according to the Valve Academic Research Consortium.[10] Bleedings were categorized into (1) life-threatening or disabling (fatal or bleeding in critical area or bleeding causing hypovolemic shock requiring surgery or drop of hemoglobin of ≥5 g/dL per transfusion of ≥4 red blood cell units), (2) major bleeding (drop of hemoglobin of ≥3 g/dL per transfusion of 2–3 red blood cell units), and (3) minor bleeding (any bleeding worthy of clinical mention).

Intracranial Hemorrhage. Definition of asymptomatic and symptomatic ICH follows the Safe Implementation of Thrombolysis criteria. Symptomatic ICH was defined as any local or remote hemorrhage that lead to a neurological deterioration of ≥4 points on NIHSS. If NIHSS was not sufficiently documented, deterioration was estimated by evaluating the description of neurological deterioration in the patient's discharge letter.

Statistical Analysis

Data are presented as frequencies (percentages) for categorical variables. Age and NIHSS are shown as medians (interquartile range, Q1–Q3) because data were not evenly distributed. Results are displayed for the whole group and stratified for the defined subgroups as recent versus nonrecent and major versus minor surgery. Rate of SSH, ICH, other bleedings, and mortality was calculated as rate of all 134 patients. The association of time of surgery (recent versus nonrecent) and type of surgery (major versus minor) with outcomes was determined using logistic regression. In addition, a high-risk group comprising patients with surgeries that were both major and recent were compared with all other patients. Odds ratios (ORs) of all subgroup analyses were adjusted for age and sex. OR of recent versus nonrecent surgery was additionally adjusted for type of surgery, OR of major versus minor surgery for time of surgery. Both crude and adjusted OR (ORadj) are presented with 95% confidence intervals. Statistical significance was determined at an [alpha]-level of 0.05 (2 tailed). Analyses were performed with IBM SPSS Statistics 22 (SPSS Inc, Chicago, IL).