Weight Approximation in Stroke Before Thrombolysis
The WAIST-Study: A Prospective Observational “Dose-Finding” Study
Background and Purpose—Because of the narrow therapeutic range for thrombolysis in stroke, accurate weight-based dosing is essential for efficacy and safety. Stroke patients are frequently incapable to communicate their correct body weight (BW). Thus, dosing is often based on BW estimation, which may lead to dosing errors. The aim of our study was to evaluate availability of BW information, accuracy of estimations, and final dosing of Alteplase (tissue plasminogen activator [tPA]) in a routine clinical setting.
Methods—A total of 109 consecutive intravenous thrombolysis patients were prospectively included in the study. Recruitment concluded with 100 complete data sets. Before therapy, BW was estimated independently by 2 physicians, 2 emergency nurses, and a neuroradiological technical assistant. Patients were weighed, and anthropometric measurements for BW approximation were taken. Dosing errors were assessed. Clinical outcome was evaluated at 90 days.
Results—Of 109 patients, 55 (50.5%) were unable to provide information on their BW. Of those, 11 (20%) were accompanied by relatives able to give BW information. For all patients, estimation errors rates ranged from 20.8% (patient’s own estimation) up to 38.2% (treating physician) and 42.2% (emergency nurse). Finally, 29 patients received an Alteplase dosage diverging >10% from the optimal dose. Twelve were under- and 17 overdosed. Underdosage was an independent predictor for worse outcome in multivariate analysis.
Conclusions—Our study shows that reliable BW data are missing for a majority of intravenous thrombolysis patients. Measuring BW before administering Alteplase remains challenging. Given dosing errors in one-third of patients and the observed impact on outcome, standardized weighing before thrombolysis should be considered.
Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifier: NCT01006434.
Thrombolysis using Alteplase (tissue plasminogen activator [tPA]) is effective in the treatment of acute ischemic stroke (AIS). Current guidelines in Western countries recommend a tPA dose of 0.9 mg/kg, up to a maximum dose of 90 mg, for patients weighing >100 kg.1 However, larger dose-finding tPA trials for intravenous thrombolysis (IVT) in AIS are missing. Based on results from research on myocardial infarction, only a few open-label studies with low case rates were initiated to evaluate the optimal dose for tPA in cerebral ischemia.2–4 These studies suggested a narrow therapeutic range, with decreased efficacy in lower dosages and an increased risk for thrombolysis related intracerebral hemorrhage (ICH) at doses of >.95 mg/kg. The ECASS-1 trial, which used a dosage of 1.1 mg/kg tPA, showed significantly higher rate of large parenchymal hemorrhages compared to trials using 0.9 mg/kg.5 Therefore, accurate dosing is crucial. In the acute phase, 2 aspects complicate tPA dosing in AIS. First, unlike in other diseases, many stroke patients are unable to communicate information on their body weight (BW) because of their stroke symptoms (eg, aphasia, decreased consciousness). In addition, motor symptoms prohibit easy weighing procedures in many patients. Second, the ultraearly and narrow time window for treatment does not allow time loss to weigh each patient in the emergency situation. Therefore, routinely, the attending physician must make a visual estimation of the patient’s BW. This may be inaccurate6–11 and may cause dosing errors, which has been shown for other weight-based emergency medication.12,13 There are few data available on tPA-dosing errors in stroke patients.14 The aim of our study was to prospectively evaluate availability of BW information, accuracy of estimations, and final dosing of tPA in a routine clinical setting. Therefore, we evaluated different sources of BW estimations and also compared visual estimation with recently proposed anthropometric measurements for BW approximation. Finally, impact of dosing errors on safety and efficacy were analyzed.
The WAIST-Study is a prospective observational monocenter study. All patients receiving IVT for acute ischemic stroke (AIS) at our institution (University Hospital Erlangen, Germany) between April 2008 and February 2009 were prospectively included in the study. Enrollment into this first part of the study concluded as prespecified after 100 complete data sets were achieved (109 patients). The study was approved by the local ethics committee and is registered with http://www.clinicaltrials.gov (Unique identifier: NCT01006434).
Visual BW Estimation and tPA Dosing
Before thrombolysis was initiated, BW of all patients was visually estimated independently by 2 physicians, 2 emergency nurses, and the neuroradiological technician. Subsequently, the patient and, if present, relatives were asked for information or estimations on the BW. Then, routine clinical practice was carried out to determine the correct dose tPA by either reaching consensus among the entire stroke emergency team in case of missing BW information or by dosing according to the known BW. For analysis, misdosing was defined as >10% divergence above (overdosing) or below (underdosing) the calculated optimal dose based on the measured BW. Estimation errors were defined accordingly as ±10% of the actual weight.14,15
Weighing and Anthropometric Measurements
Data Collection and Evaluation of Outcome
Standard baseline demographic and stroke related data, as well as treatment specifics and imaging information, were collected. Additionally, we recorded for each patient the actual weight, the calculated optimal dosage, the applied dosage, and the weight estimations of the treating physicians, nurses, and neuroradiological technicians. Furthermore, it was documented whether the patient or relatives were able to give information on the BW. All patients were treated and monitored on our stroke unit according to European guidelines.1 NIHSS score was recorded by a stroke neurologist certified for NIHSS evaluation at presentation and 3 times per day throughout the hospital stay. Outcome at 90 days was assessed by a physician blinded to the BW and dosage data using the modified Rankin scale (mRS) by performing a semistructured interview either in person or by telephone. All raters were certified for the NIHSS and the mRS. Outcome was dichotomized into favorable (mRS, 0 to 1) versus unfavorable outcome (mRS, 2 to 6) and independent outcome (mRS, 0 to 2) versus dependency and death (mRS, 3 to 6). Symptomatic and asymptomatic hemorrhages were defined using the ECASS-3 criteria.16
Statistical analyses were performed using the SPSS software package (version 14.0; SSPS Inc). All data were tested for normality and are given as median and range or mean and SD as appropriate. Categorical variables are presented as frequencies and percentages. In group comparisons, the statistical significance of differences among patients with over- or underdosage compared to all other patients was tested with the χ2 test. To identify independent predictors for favorable (mRS, 0 to 1) and independent (mRS, 0 to 2) outcome, univariate analysis was carried out, and parameters showing at least a trend (P<0.2) were entered into multivariate stepwise backward logistic regression analysis.
Baseline Characteristics and Treatment
During an 11-month period, 109 patients with acute ischemic stroke who received IVT at our dedicated neurological emergency room were enrolled into this study. Nine patients had to be excluded from some analyses because of incomplete data (malfunction of bed scale in 3 patients, 4 patients with medical instability prohibiting weighing on the intensive care unit, 1 patient died before weighing, and 1 patient was transferred to the cardiology department directly after thrombolysis). However, these patients were included into the analysis wherever possible to prevent bias. Patient characteristics are presented in Table 1. Median age was 71 years (interquartile range, 60 to 79.5) and 61% were male. The median NIHSS at admission was 9 (interquartile range, 4.5 to 13). At day 90, the median mRS was 3, with 36 (33%) patients achieving a favorable and 52 (48%) an independent outcome. Eight (7.3%) asymptomatic and 5 (4.6%) symptomatic ICHs occurred. The mean BW was 77.5 kg (SD: 16.7 kg), and the mean tPA dosage was 0.904 mg/kg BW (SD: 0.101 mg/kg). Forty-seven (43.1%) patients were weighed in a standing position and 55 (50.5%) patients using the bed scale. No significant differences in baseline characteristics between correctly dosed or under- and overdosed patients were noted, except that overdosed patients had a lower BW and a slightly higher baseline NIHSS score.
BW Information and Estimations
Of the 109 treated patients, 55 (50.5%) were unable to provide information about their BW. Eleven (20%) of them were accompanied by relatives who were able to give such information (Table 2). BW estimation error rates are given in Table 3. They ranged from 20.8% (11.3% overestimation [OE]; 9.4% underestimation [UE]) for the patients themselves, up to 44.4% (29.3% OE; 15.2% UE) for the neuroradiological technicians. Error rates were reduced from 42.2% to 33.3% (20.6% OE; 12.7% UE) and from 38.2% to 34.7% (20.8% OE; 12.8% UE) once the mean values of 2 nurses or 2 physicians were taken, respectively. Using the anthropometric BW approximation formula,15 the overestimation error and underestimation error were 17.9% and 2.1% (Table 3).
Dosage, Outcome, and ICH Rate
Comparing the actual applied dosage and the optimal dosage (based on the measured BW), 33% of patients received an tPA dosage diverging >10% from the optimal dose. Of those, 17 (17%) patients were over- and 16 (16%) underdosed. Nine (8.3%) of the 109 patients weighed >100 kg and received the maximum dose of 90 mg tPA. Taking this into account, the corrected number of patients who received an underdosage of >10% declines to 12 (12%), corresponding to an overall misdosing in 29% of patients (Table 4).
Group comparisons of outcome and safety for patients receiving the correct dose and under- and overdosed patients are shown in Table 5. Fewer patients in the underdosed group reached favorable and independent outcomes. Overdosage had neither impact on clinical outcome nor on safety of therapy. No symptomatic ICH and only 1 asymptomatic ICH occurred within the group of overdosed patients; thus, no significant increase of ICH could be observed compared to the correctly dosed patients (Table 5).
In univariate regression analysis, NIHSS score at baseline, age, and underdosage were relevant factors for favorable and independent outcome (data not shown) and were thus entered into multivariate regression models. In multivariate analysis, only NIHSS score on admission (odds ratio [OR]: 1.2; 95% confidence interval [CI]: 1.09 to 1.33; P<0.001) remained as an independent predictor for an unfavorable outcome (mRS, 2 to 6), whereas underdosage showed a nonsignificant trend (OR: 4.16; CI: 0.78 to 22.23; P=0.09; Table 6). Regarding dependency and death (mRS, 3 to 6) underdosage (OR: 5.87; CI: 1.26 to 27.34; P=0.024) and a higher NIHSS on admission (OR: 1.24; CI: 1.12 to 1.37; P<0.001) independently predicted a worse outcome (Table 6).
Alteplase in AIS has a low therapeutic range requiring a strict weight-based dosing regimen. Especially in emergency situations, there are often neither the means nor the time to weigh each patient before administering medication. Thus, visual BW estimation has become the standard in these cases. There are only few data on availability and accuracy of BW estimation, accuracy of tPA dosing, and outcome after dosing errors in ischemic stroke patients. This prospective monocenter study assessed those issues in a clinical routine setting.
BW estimation in children17–19 and adults as a fast alternative to weighing has been explored in other studies before. These studies range from patients in emergency departments7,8,10,11,20–22 and in preoperative situations6 to healthy volunteers.9 Estimation errors (EEs) up to 45% for medical professionals and 10% for the patients’ own estimation have been reported. However, within the cited studies, neither the patient’s position (eg, supine, standing) nor the number and experience of the participating estimating personnel was consistent. Whereas some trials included emergency patients, others excluded medically unstable patients or even examined healthy volunteers. Therefore, comparability to our study is limited. The largest trial to date for emergency patients prospectively examined accuracy of weight estimation in 1137 patients in a supine position and described an EE of 41% for physicians, 22% for nurses, and 9% for the patients.11 Although the numbers reported for estimations by the medical professionals in the literature are comparable to our trial, percentages for the patient’s own estimation seem to be more favorable in nonstroke populations. This highlights one of the specific problems encountered in an emergency stroke population. Many stroke patients are unable to communicate information on their BW because of their neurological symptoms (eg, aphasia, decreased consciousness). In our study, more than half of the patients could not provide information on their BW. In some cases, aphasic patients might even provide wrong BW information, potentially leading to a higher rate of EE for the patient’s own estimation observed in our study. Further complicating matters, often motor symptoms prohibit fast weighing procedures.
There are very few comparable data on weight estimations in stroke patients. In the study by Lorenz et al, 178 patients were included and again EEs were in the same range as reported before.15 However, in contrast to our study, only 6.2% of their patients received IVT, 20.8% experienced a transient ischemic attack, and 9.6% had stroke mimics. No data on the NIHSS are given, but 77.5% of the patients were able to stand on admission. Our patient collective and approach were different. We focused on the hyperacute and narrow time window for stroke treatment, which, together with the treatment decision itself, puts pressure on the physician. In addition, inclusion of IVT patients only allowed us to examine the dosing procedure in the affected patient population itself, analyze the finally applied dosage, and detect dosage errors resulting from the dose finding process routinely used in clinical practice. After all information was used to calculate the optimal dose, still almost one-third of patients received a dosage of tPA diverging >10% from the calculated optimal dose. This number is higher than the dosing error rate of 12% reported by a retrospective stroke survey including data from several different international cohorts.14 However, in this survey, it was not clear how BW information was obtained. Of the 1205 included patients, only 769 (64%) had weight and dosage data, and 385 of these (50%) received the optimal dose. Thus, even the authors of this survey wondered whether a high number of patients were weighed or BW was calculated retrospectively for the survey using the actually applied tPA dosage.
Knowing that this pilot study was not powered to conclusively answer these issues, 2 hypotheses were tested using the obtained dosing information: first, whether receiving too little tPA reduces the chance for a favorable or independent outcome; and second, whether overdosed patient exhibit a higher risk for hemorrhagic complications. Regarding outcome, underdosage was a strong predictor for worse outcome. In multivariate analysis, only NIHSS on admission and being underdosed were independent factors for dependency and death (mRS, 3 to 6), and underdosage also showed a trend toward predicting unfavorable outcome (mRS, 2 to 6). No hemorrhages were seen in patients receiving too much tPA. These results are in contrast to the data reported in the retrospective survey by Messe et al,14 who reported unchanged efficacy in underdosed and an elevated risk for asymptomatic, although not for symptomatic hemorrhages in overdosed patients. However, missing data on two-thirds of patients in this survey may have confounded outcome analysis.
Another approach of our study was to use the anthropometric BW approximation formula introduced by Lorenz et al in 2007.15 The observed EE rate of 20% in our cohort was much higher than described in the original publication (6.2%),15 suggesting that this formula, which uses body constitution (height and waist and hip circumference) must be validated separately for different populations. Interestingly, with 2.1%, the underestimation rate was very low, whereas overestimation occurred in 17.9% when using the formula. Although more accurate than the visual estimations of our clinical personnel, using the formula alone would have caused a selective tPA overdosing in our patient collective. It must noted that performing anthropometric measurements was at least as much of an effort in our patient collective with mainly bedridden patients as actually weighing the patients. However, the former may represent an alternative if no scale is available.
Our study has some limitations, most notably the low case number; therefore, the results must be interpreted with caution. It is clearly underpowered to allow definite conclusions on efficacy and safety of IVT in misdosed patients. However, even with a low case number, this pilot study demonstrated a significant effect of underdosing of tPA on clinical outcome. It further allows effect estimation for future research. Because of the monocenter approach results might be influenced by local aspects such as the emergency room setting and experience of staff. Another potential shortcoming is the fact that patients were weighed within 24 hours after IVT, which might have caused some inaccuracy.
In conclusion, our study shows that reliable information on the patient’s BW is missing in a majority of stroke patients in the acute setting of thrombolytic therapy. Furthermore, visual estimation is inaccurate in a substantial proportion of patients, and even using all available information, the dose finding procedure commonly used in routine clinical practice leads to misdosing in approximately one-third of all patients. Based on these data, we now continued enrollment into the follow-up WAIST-II study, which is planned to be expanded to a multicenter study. The bed scale was relocated to the emergency room, and the study protocol allows weighing of the patient immediately after starting thrombolysis and adapting the tPA dosage whenever indicated. However, initial experience shows that weighing of IVT patients within the 1-hour infusion time remains challenging, and a proportion of patients complete thrombolysis before weighing can be carried out. WAIST-II will (1) show the degree to which tPA under- and overdosing still occur using such a protocol; and (2) examine the clinical impact of dosing errors in a larger population assembled from different stroke centers.
Sources of Funding
No funding was involved in the present study.
P.D.S., M.K., H.B.H., R.K., and S.S. received travel grants from Boehringer Ingelheim. S.S. and P.D.S. are members of the advisory board and received speaker honoraria from Boehringer Ingelheim, the manufacturer of rt-PA.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
- Received January 5, 2010.
- Accepted February 5, 2010.
Brott TG, Haley EC Jr, Levy DE, Barsan W, Broderick J, Sheppard GL, Spilker J, Kongable GL, Massey S, Reed R, Marler JR. Urgent therapy for stroke. Part I. Pilot study of tissue plasminogen activator administered within 90 minutes. Stroke. 1992; 23: 632–640.
Haley EC Jr, Levy DE, Brott TG, Sheppard GL, Wong MC, Kongable GL, Torner JC, Marler JR. Urgent therapy for stroke. Part II. Pilot study of tissue plasminogen activator administered 91–180 minutes from onset. Stroke. 1992; 23: 641–645.
Levy DE, Brott TG, Haley EC Jr, Marler JR, Sheppard GL, Barsan W, Broderick JP. Factors related to intracranial hematoma formation in patients receiving tissue-type plasminogen activator for acute ischemic stroke. Stroke. 1994; 25: 291–297.
Hacke W, Kaste M, Fieschi C, Toni D, Lesaffre E, von Kummer R, Boysen G, Bluhmki E, Hoxter G, Mahagne MH, Hennerici M. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The European Cooperative Acute Stroke Study (ECASS). JAMA. 1995; 274: 1017–1025.
Cubison TC, Gilbert PM. So much for percentage, but what about the weight? Emerg Med J. 2005; 22: 643–645.
Martin DR, Soria DM, Brown CG, Pepe PE, Gonzalez E, Jastremski M, Stueven H, Cummins RO. Agreement between paramedic-estimated weights and subsequent hospital measurements in adults with out-of-hospital cardiac arrest. Prehosp Disaster Med. 1994; 9: 54–56;discussion 57.
Freedman JE, Becker RC, Adams JE, Borzak S, Jesse RL, Newby LK, O'Gara P, Pezzullo JC, Kerber R, Coleman B, Broderick J, Yasuda S, Cannon C. Medication errors in acute cardiac care: an American Heart Association Scientific Statement from the Council on Clinical Cardiology Subcommittee on Acute Cardiac Care, Council on Cardiopulmonary and Critical Care, Council on Cardiovascular Nursing, and Council on Stroke. Circulation. 2002; 106: 2623–2629.
Lorenz MW, Graf M, Henke C, Hermans M, Ziemann U, Sitzer M, Foerch C. Anthropometric approximation of body weight in unresponsive stroke patients. J Neurol Neurosurg Psychiatry. 2007; 78: 1331–1336.
Sanchez LD, Imperato J, Shapiro N. Weight estimation by emergency department personnel. Acad Emerg Med. 2004; 11: 546.