Clinical Biochemistry 48 (2015) 483–488
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Clinical
Quality indicators for critical international normalized ratio measurements in a hemostatic laboratory Shabneez Hussain a, Bushra Moiz a,⁎, Bushra Afaq b a b
Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University Hospital, Karachi, Pakistan Coagulation, Clinical Laboratory, Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University Hospital, Karachi, Pakistan
a r t i c l e
i n f o
Article history: Received 6 November 2014 Received in revised form 10 December 2014 Accepted 21 December 2014 Available online 3 January 2015 Keywords: Critical international normalized ratio Quality indicators Warfarin Communication Oral anticoagulants HAS-BLED score Turnaround time
a b s t r a c t Objective: This study aimed at reviewing the quality indicators for reporting critical international normalized ratio (INR) in a coagulation laboratory. Design and methods: This is a retrospective study conducted at Aga Khan University Hospital, hemostatic laboratory from February 2010 till January 2011. Critical INR was defined as ≥5.0. All critical INRs were rechecked and results were communicated to the doctor or patient. Two quality indicators monitored were % of results communicated to the patient/doctor and % of results that remained critical after re-testing. Results: During the study period, a total of 59,980 INRs were reported. Of these 376 or 0.6% were critical. Successful communication of critical results to the doctor or patient was achieved in 275/376 (73.1%). Overall 353 or 94% (343 initial and 10 re-draw) samples had critical INR on repetition. Twenty five patients of the 240 warfarinized patients with critical INR had mild bleeding. No life threatening bleeding was observed in any patient. Conclusion: We observed poor communication for notifying critical INR results during the study period. Routine repeat analysis of critical INR did not alter results in majority of samples. The study assisted in improving communication in subsequent years. Further work is needed to establish evidence based upper and lower cutoff of critical INR. Effect of replicate analysis on turnaround time and accuracy of results needs evaluation. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Novel oral anticoagulants have provided the clinicians with broader choice for anticoagulation in thrombo-embolic conditions. However, vitamin K antagonists such as warfarin, acenocoumarol and phenprocoumon are still used as the drugs of choice in developing countries because of their cost effectiveness and physician preferences. Warfarin inhibits vitamin K epoxide reductase thereby decreasing the synthesis of vitamin K dependent clotting factors (II, VII, IX, X) and natural anticoagulants (protein C and S). It requires laboratory monitoring which is routinely performed by measuring international normalized ratio (INR). For optimum management, INR is maintained within a narrow therapeutic range of 2–3 [1,2]. Similar to all other anticoagulants, overdosing with
Abbreviations: INR, International normalized ratio; ISO, International Organization for Standardization; JCIA, Joint Commission International Accreditation; WHO, World Health Organization; AKUH, Aga Khan University and Hospital; ILMS, Integrated Laboratory Management System; PT, Prothrombin time; APTT, Activated partial thromboplastin time; TAT, Turnaround time. ⁎ Corresponding author at: Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University and Hospital, Stadium Road, P.O. Box 3500, Karachi 74800, Pakistan. Fax: +92 21 34934294, +92 21 34932095. E-mail addresses: [email protected] (S. Hussain), [email protected] (B. Moiz), [email protected] (B. Afaq).
warfarin can lead to bleeding complications which has an incidence of 10 to 17% per year for all events. Major bleeding is described as fatal or intracranial (documented by imaging), ocular (with blindness), retroperitoneal or intra-articular bleeding [3]. Another criterion is severe bleeding causing a fall of 20 g/L (1.24 mmol/L) hemoglobin or more or requiring ≥2 units of whole/red blood cells [3,4]. The annual incidence of major bleeding is 2 to 5% with 0.5 to 1.0% fatal bleeding [5] and 0.2 to 0.4% intracranial bleeding [6,7]. In 1972, Lundberg was the first to describe the communication of critical laboratory tests results to the physicians for possible intervention [8]. Communicating critical results is usually an established policy for quality practice by many accrediting agencies like the International Organization for Standardization (ISO), the Joint Commission International Accreditation (JCIA) and the World Health Organization (WHO). The Aga Khan University and Hospital (AKUH) is a tertiary care hospital situated in Southern Pakistan. It was accredited by ISO 9001 in June 2000 and by JCIA in September 2006. The system of notification of critical results was established in 2007 in the hospital which was switched to the computerized Integrated Laboratory Management System (ILMS) in 2009. Among other critical results, an INR of five or above was considered reportable to health care providers [9,10]. This cutoff was chosen based on published literature and physicians' experiences with warfarin associated bleeding. It is important for a hemostatic laboratory to report
http://dx.doi.org/10.1016/j.clinbiochem.2014.12.019 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
484
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
high INR so that warfarin therapy can be temporarily discontinued. At the same time, it is imperative that laboratory measurement is repeated to observe if a critical INR becomes non-critical on repetition. This study is aimed at reviewing the performance of coagulation laboratory for the notification of critical INR and agreement of INR measurements on repeat test samples. Secondary objective was to define a cutoff for critical INR results reporting. The review was expected to provide results for further improving clinical services of the laboratory. Materials and methods Setting and critical result reporting The AKUH is a 700 bedded tertiary care academic institute with facilities for large oncology and trauma units as well as a fully functional bone marrow transplant unit. The hemostatic laboratory at AKUH is responsible for testing the blood samples of all admitted patients. Additionally, the laboratory acts as a national referral center and receives specimens through 192 blood collection centers. Approximate daily test volumes are 250 prothrombin time (PT)/INR and 150 activated partial thromboplastin time (APTT) samples. Our lab reports INR with every PT test result by default. Our system of critical reporting was established in 2007 as the hospital was accredited with JCIA. Initially the system did not alert either clinical or laboratory personnel for critical results. In 2009, our information structure was upgraded and the computerized system notified physicians for critical results for admitted patients. However, for outpatients, the laboratory technologist/resident communicated the results to the concerned doctor or patient.
was performed on Sysmex ®cs2000i using Dade® Innovin® reagent with an ISI of 1.02. INR was computed by the instrument by calculating [patient PT / control PT]ISI. A critical INR was defined as 5.0 or above. As per our laboratory policy, all PT and INR results in critical range were repeated. If more than 10% discordance was observed between two results then a fresh sample was requested. Delta check was possible only for in-patients. For out-patients, every lab visit was coded individually and therefore previous test results of the same patient could not be retrieved from the system. All verified critical results for admitted patients were communicated to the health care providers within an hour of analysis and read back policy was followed. Outside referrals were communicated to either doctors or patients themselves. A detailed history of the patient was taken by an on call doctor in the laboratory to evaluate the cause of high INR (see Fig. 2). If the patient was found taking warfarin then he was advised to hold the next dose of warfarin and communicate with his physician on urgent basis. In case of laboratory failure to communicate, each patient was given two additional telephone calls within 24 h. In the absence of a response, the information was entered in the computerized system and the patient was considered unapproachable. Evaluation of performance indicators To review the lab performance, we calculated the percentage of successfully communicated critical results of INR to patients/care givers, the percentage of repeated results that met the criteria for critical results and the percentage of repeated results that significantly drifted from initial test result. These performance indicators were adopted from the College of American Pathologists [11] and were calculated as follows:
Study design This was a retrospective study that was conducted from February 2010 to January 2011. Data collection Overall laboratory process of notification of high INR is detailed in Fig. 1. Briefly, blood samples were collected in 3.2% Na citrate and PT
1. % of successful communication of INR = number of patients who were notified of results / total number of patients with critical INR tests × 100 2. % of critical INR that stays critical = number of critical INR re-tested and stayed critical / total number of critical INR tests × 100 3. % of critical INR that became non-critical = number of critical INR re-tested and changed to non-critical INR / total number of critical INR tests × 100.
Fig. 1. Method of communicating and documenting critical INR as established in 2009.
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
485
Fig. 2. Questionnaire for reporting critical INR (≥5.0).
Evaluation of other variables for patients on warfarin therapy During the study period, various parameters of warfarinized patients were also studied such as indications, adverse effects, duration of treatment and frequency of INR monitoring. Special emphasis was given for the site and severity of bleeding events at the time of reporting critical INR. HAS-BLED score was computed for bleeding patients to assess the risk of bleeding [12]. The scoring was done by summation of one point each for history of hypertension, stroke, liver, renal diseases, labile INR, age N 65 years, antiplatelet agents and alcohol consumption.
ranged from 5 to 15 (mean ± SD; 6.8 ± 1.7) with a male to female ratio of 1:1.4. There were ten pediatric patients with age b 15 years, 140 (37%) were between 15 and 40 years and 226 or 60% patients were above 50 years. Of the 376 critical INR, 240 (63.6%) patients were taking warfarin while eight (2%) had other coagulopathies such as liver dysfunction or vitamin K deficiency. The cause of critical INR remained undetermined in 128 (34.3%) patients due to communication failure (n = 101) or patients' unawareness of their disease/drugs intake (n = 27).
Statistical analysis
Quality indicators for laboratory performance
All data was entered into SPSS version 21.0 (IBM, Armonk, NY, USA). Descriptive data was given as number (%) and frequency for qualitative variables and mean ± standard deviation for quantitative variables. Frequency and percentages were calculated to check the association between bleeding and non-bleeding patients with different parameters, and fisher exact test or chi square test was applied.
Of the 376 critical INR, 275 results were communicated to patients/ health professionals constituting 73% of total critical INR. Reason for communication failure in remaining critical results (n = 101 or 26.9%) was absent or incorrect contact number. As per policy, every abnormal test was repeated on the same sample. Three hundred forty-three (91.2%) results remained unchanged in following retesting. New samples were requested from 33 patients with 12 refusing for redraw. Of the 21 fresh samples obtained, INR changed from a mean critical value of 7.1 ± 0.95 to a non-critical mean of 1.4 ± 0.7 in 11 patients while remaining critical (from an original mean of 7.2 ± 0.94 to 6.6 ± 0.75) in 10 patients. Thus, overall 353 or 94% (343 initial and 10 re-draw) samples had critical INR on repetition. A further analysis showed that 10 of 11 samples that became non critical (b5.0) were received from various blood collecting centers as frozen samples. On investigation, it was found that loss of sample integrity during transport and inappropriate collection technique was responsible for spuriously raised INR in samples received from collection centers. Hence pre-analytic error was the reason behind this high level of disagreement.
Ethical approval The study was approved by the ethical review committee of The Aga Khan University (#2526-Obs-ERC-13). Results Demographics During the study period, 59,980 total and 376 (or 0.6%) critical (≥ 5.0) INRs were reported on outpatient samples. Critical INR
486
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
Characteristics of warfarinized patients Two hundred and forty of 376 patients with critical INR were taking warfarin. There were 145 females and 95 males with a mean age of 55.7 years ± 18.1 and 89 patients (37%) above the age of 65 years. Seventy four (31%) patients were on lifelong treatment while 77 (32%) and 17 patients were taking warfarin respectively for 0–6 and 6–12 months. Seventy two patients (30%) could not recall the dates of warfarin initiation. Frequency of INR measurement was weekly or biweekly in 73 (30%) and monthly in 49 patients (20%) while no set pattern was observed in the remainder. Valvuloplasty (n = 53 or 22%) was the commonest indication of warfarin intake followed by deep venous thrombosis (n = 47 or 20%) and cerebrovascular accident (n = 42 or 18%). Less common indications included coronary by-pass surgeries/ angioplasty (n = 27 or 11%), pace makers/arrhythmia (n = 17 or 7%), myocardial infarction (n = 11 or 5%) and polycythemia (n = 1) while indications were not known to 42 patients. Majority of the patients (n = 63 or 26%) were on anti-hypertensive medications while 32 or 13% were taking aspirin. Sixty four patients (27%) were taking at least two or more medications beside warfarin at the time of INR testing. Warfarin toxicity and critical INR Of the 240 patients on warfarin, adverse effects were observed in 43 (18%) including bleeding (n = 24), headache (n = 13), diarrhea (n = 5) and combined bleeding and diarrhea (n = 1). Gingival bleeding was the commonest complaint observed in ten patients. Others were bruising (n = 4), hemoptysis (n = 2), hematoma (n = 2), gastrointestinal (n = 1), menorrhagia (n = 4), post-operative (n = 1) and hemarthrosis (n = 1). Majority of hemorrhagic patients (n = 16) were females. Nine patients were on extended warfarin therapy and six and four were taking warfarin secondary to valvuloplasty and CVA respectively. Only three of 25 hemorrhagic patients were taking aspirin. No major or life threatening bleeding was observed. INR was ≥10.0 in three patients and it ranged from 5.0 to ≤ 10.0 in 22 patients [Fig. 3]. Majority (n = 10) of the bleeding patients had INR ranging from 5.0 to ≤ 6.0. HAS-BLED scores were 0, 1, 2 and 3 respectively in 10, 5, 6 and 3 patients with bleeding [Fig. 4]. Table 1 gives comparative analysis of various parameters of two groups of patients with and without bleeding. History of myocardial infarction and INR were statistically significant for bleeding. Characteristics of patients with critical INR due to other causes (n = 136) Though our data focuses on the warfarinized group, information for 136 patients with critical INR due to other causes was also available.
Fig. 4. Comparison of HASBLED score in warfarinized [n = 240] patients with [n = 25] and without bleeding [n = 215].
These were 73 (57%) females and 55 (43%) males. The mean age was 50 ± 17 years and the mean critical INR was 7.1 ± 2.1. No contact was possible in 101/136 cases due to incorrect or no contact number. Of the remaining 35, a large group was ignorant of disease/drug intake (n = 27) while eight patients suffered from coagulopathies. Of those with coagulopathies (n = 8), two patients had bruises, one patient (3 years of age) bled after circumcision and one patient had gastrointestinal bleeding. Hence adverse effects of bleeding were observed in 4/8 or 50% patients emphasizing the significance of communicating critical INR even in non-warfarinized patients. Discussion We observed 0.6% critical INR results during the study period. Successful communication to outside referrals was 73% while 93% test results stayed critical after repeat measurements. We had a high communication failure of 37% for out-patients. Threshold for critical INR In 2007, we identified critical values for various chemistry and hematology tests based on published literature. A critical INR was set at 5.0 or above in our laboratory at which only 10% of the warfarinized patients were minimally hemorrhagic. Their INR ranged from 5 to 15 and HAS-BLED scoring was 0 to 3. No life threatening episodes or major bleeding events were observed in patients communicated to. Most
Fig. 3. Comparison of critical INR in warfarinized [n = 240] patients with [n = 25] and without bleeding [n = 215].
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488 Table 1 Base line characteristics of warfarinized patients [n = 240] with and without bleeding having critical INR. Descriptive values are mean ± 1SD.
487
consequence. According to JCIA, critical results should be informed to clinician who can take immediate action [19].
Variable
Non-bleeding n = 215
Bleeding n = 25
Total n = 240
p-Value
Testing frozen samples and INR reporting
Age N 65 years n (%) Females n (%) Aspirin intake Hypertension Diabetes mellitus Previous stroke Previous myocardial infarction Previous valvuloplasty Previous CABG Atrial fibrillation Monthly INR monitoring Time since warfarin intake in months (mean ± SD) INR (mean ± SD) HAS-BLED score (mean ± SD)
81 (37.7) 128 (59.5) 29 (13.5) 54 (25.1) 12 (5.6) 37 (17) 7 (3.3) 47 (22) 24 (11) 16 (7.4) 118 (54.8) 4.6 ± 2.7
8 (32) 17 (68) 3 (13) 9 (36) 2 (8) 5 (20) 4 (16) 6 (24) 3 (12) 1 (4) 14 (56) 4.0 ± 2.4
89 145 32 63 14 42 11 53 27 17 132 (55) 4.6 ± 2.6
0.774 0.413 0.836 0.242 1.000 0.728 0.015⁎ 0.807 0.728 1.000 0.843 0.409
Frozen plasma samples with elevated INR received from phlebotomy centers could not be differentiated from EDTA, heparinized plasma or even from serum. When coagulation samples are received as frozen ‘plasma’ and show prolonged PT/APTT, CAP [20] recommends testing fibrinogen levels. A very low fibrinogen level below detection limit in absence of warfarin or bleeding history would imply testing on fresh sample. However, due to financial restrain, we do not perform fibrinogen assay on any sample but rather request fresh sample for repeat testing.
6.6 ± 1.2 1.0 ± 1.0
7.1 ± 2.7 1.1 ± 1.0
6.6 ± 1.5 1.0 ± 1.0
0.042⁎ 0.781
⁎ p value less than 0.05 is statistically significant.
laboratories use a critical value threshold of 4 or 5 but a wide range of 2.6 to 10 has been observed [13]. A critical value needs to be optimally chosen and must be evidence based to represent a life threatening situation for reducing mortality [14]. A lower cutoff could not be established by us and patients bleeding on lower INR would not have been captured. Repeating analysis Every critical result was repeated and 91% of our results remained unchanged. It is argued by Niu A and Deetz CO [15,16] that such repetition increases the turnaround time and cost of the testing without any added accuracy. Toll AD in 2011 [17] described differences between two test runs of 553 individual critical PT results. He found a maximum difference of 4 s in 99.4% results. He concluded that repeat testing in critical testing did not provide any advantage in coagulation setting. A recent evaluation of various tests (not including INR) in 86 laboratories showed that routine repeat analysis of automated hematology and chemistry critical values is unlikely to be useful and may adversely affect patient care [18]. Some laboratories rely on delta check to determine the need of repeating. However, in our case this was not possible as our ILMS did not allow tracking of previous test results in outpatients. In our laboratory the mean turnaround time (TAT) for reporting a verified critical result was 45–60 min. At the time, this study was done there was no stress on TAT for reporting. TAT for reporting INR on a fresh sample was 2 h after receiving the sample. Hence where a redraw was required for INR confirmation, reporting would take 24 h after first sample draw. JCIA recommends an explicit time frame of 30 min between completion of test and receipt by responsible care giver. In 2013, ILMS was upgraded to capture this time interval and measured quarterly as a quality indicator of the lab. In 2012, this time line was achieved in 47% of the cases [data not shown] and improved to 97% in 2013. Communication failure This was a major barrier in our situation. Contact numbers provided were either incorrect or patients did not receive calls. Similarly physicians' contact numbers are unknown to patients themselves who visit laboratory for their testing. To handle this situation, it was suggested that patients should be warned at the time of blood draw that they might receive a call from the laboratory in case of critical results. E-mails/fax could be suitable alternative tools of communication. In Pakistan, notification alerts are directed mainly towards patients or his relatives and rarely to care givers. Our experience in communicating results to patients was very difficult as they might not understand the
One critical value for all We had one critical INR value of N 5.0 for all types of patients. In certain situations what is “life-threatening” for one group may not be for all others and vice versa. Working with clinicians is imperative to understand the processes in “unique” locations e.g. an INR of N 1.5 may be critical for a patient with anticipated bleeding such as an invasive procedure [21]. Conclusions We observed poor communication for notifying critical INR results during the study period. Routine repeat analysis of critical INR did not alter results in majority of the samples. The study assisted in improving communication in subsequent years. Further work is needed to establish evidence based upper and lower cutoff of critical INR. Effect of replicate analysis on TAT and accuracy of results needs evaluation. Conflict of interest The authors declare that there are no conflicts of interest. Authors information SH is a fifth year resident in the Hematology Department at the Aga Khan University and Hospital. BM is an associate professor and consultant hematologist at the Aga Khan University and Hospital. BR is the assistant manager in the Coagulation section of the Clinical Laboratory. Author contributions SH analyzed the data and wrote the original draft of the paper. BR analyzed the data and BM provided the original idea for conducting this audit, guided the work, reviewed the paper, analyzed the data and wrote some of the manuscript. All authors were involved in revising and approving the manuscript. Acknowledgments We thank Dr Rabia Rafiq, Mehwish Arshad and Subuhi Raza for collecting the data. We also thank Mr. Ahmad Raheem for analyzing the data. References [1] ICSH/ICTH recommendations for reporting prothrombin time in oral anticoagulant control. International Committee for Standardization in Haematology and International Committee on Thrombosis and Haemostasis. J Clin Pathol Feb 1985;38(2): 133–4 [PubMed PMID: 3968214. Epub 1985/02/01. eng]. [2] Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians
488
[3]
[4]
[5]
[6]
[7]
[8] [9]
[10]
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488 Evidence-Based Clinical Practice Guidelines (8th Edition). Chest Jun 2008;133(6 Suppl.):160S–98S [PubMed PMID: 18574265. Epub 2008/07/24. eng]. Schulman S, Kearon C. Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost Apr 2005;3(4): 692–4 [PubMed PMID: 15842354]. Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D'Angelo A, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet Aug 17 1996;348(9025):423–8 [PubMed PMID: 8709780. Epub 1996/08/ 17. eng]. Stehle S, Kirchheiner J, Lazar A, Fuhr U. Pharmacogenetics of oral anticoagulants: a basis for dose individualization. Clin Pharmacokinet 2008;47(9):565–94 [PubMed PMID: 18698879. Epub 2008/08/14. eng]. Schulman S, Beyth RJ, Kearon C, Levine MN. Hemorrhagic complications of anticoagulant and thrombolytic treatment: American College of Chest Physicians EvidenceBased Clinical Practice Guidelines (8th Edition). Chest Jun 2008;133(6 Suppl.): 257S–98S [PubMed PMID: 18574268. Epub 2008/07/24. eng]. Palareti G, Cosmi B. Bleeding with anticoagulation therapy — who is at risk, and how best to identify such patients. Thromb Haemost Aug 2009;102(2):268–78 [PubMed PMID: 19652877. Epub 2009/08/05. eng]. Lundberg GD. When to panic over abnormal values. MLO Med Lab Obs 1972;4: 47–54. Fairweather RB, Ansell J, van den Besselaar AM, Brandt JT, Bussey HI, Poller L, et al. College of American Pathologists Conference XXXI on laboratory monitoring of anticoagulant therapy: laboratory monitoring of oral anticoagulant therapy. Arch Pathol Lab Med Sep 1998;122(9):768–81 [PubMed PMID: 9740135. Epub 1998/09/18. eng]. Hirsh J, Dalen JE, Anderson DR, Poller L, Bussey H, Ansell J, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest Nov 1998;114(5 Suppl.):445S–69S [PubMed PMID: 9822057. Epub 1998/11/20. eng].
[11] 2010_qmt_catalog.pdf. [12] Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest Nov 2010;138(5):1093–100 [PubMed PMID: 20299623]. [13] Lusky K. INR practice gaps found in Q-Probes 2011 [cited 2014 30-10-14]. Available from: http://www.cap.org/apps//cap.portal?_nfpb=true&cntvwrPtlt_ actionOverride=%2Fportlets%2FcontentViewer%2Fshow&_windowLabel= cntvwrPtlt&cntvwrPtlt%7BactionForm.contentReference%7D=cap_today%2F0311% 2F0311c_inr_practice.html&_state=maximized&_pageLabel=cntvwr. [14] Doering TA, Plapp F, Crawford JM. Establishing an evidence base for critical laboratory value thresholds. Am J Clin Pathol Nov 2014;142(5):617–28 [PubMed PMID: 25319976]. [15] Deetz CO, Nolan DK, Scott MG. An examination of the usefulness of repeat testing practices in a large hospital clinical chemistry laboratory. Am J Clin Pathol Jan 2012;137(1):20–5 [PubMed PMID: 22180474]. [16] Niu A, Yan X, Wang L, Min Y, Hu C. Utility and necessity of repeat testing of critical values in the clinical chemistry laboratory. PLoS One 2013;8(11):e80663 [PubMed PMID: 24260448. Pubmed Central PMCID: 3834106]. [17] Toll AD, Liu JM, Gulati G, Behling EM, Kocher WD. Does routine repeat testing of critical values offer any advantage over single testing? Arch Pathol Lab Med Apr 2011;135(4):440–4 [PubMed PMID: 21466359]. [18] Lehman CM, Howanitz PJ, Souers R, Karcher DS. Utility of repeat testing of critical values: a Q-probes analysis of 86 clinical laboratories. Arch Pathol Lab Med Jun 2014;138(6):788–93 [PubMed PMID: 24878017]. [19] Hanna D, Griswold P, Leape LL, Bates DW. Communicating critical test results: safe practice recommendations. Jt Comm J Qual Patient Saf Feb 2005;31(2):68–80 [PubMed PMID: 15791766]. [20] Pathologists CoA. Hematology and coagulation checklistIn: Pathologists CoA, editor. 2011. p. 8. [21] Douketis JD, Bakhsh E. Perioperative management of antithrombotic therapy. Pol Arch Med Wewn Apr 2008;118(4):201–8 [PubMed PMID: 18575419].
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Clinical
Quality indicators for critical international normalized ratio measurements in a hemostatic laboratory Shabneez Hussain a, Bushra Moiz a,⁎, Bushra Afaq b a b
Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University Hospital, Karachi, Pakistan Coagulation, Clinical Laboratory, Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University Hospital, Karachi, Pakistan
a r t i c l e
i n f o
Article history: Received 6 November 2014 Received in revised form 10 December 2014 Accepted 21 December 2014 Available online 3 January 2015 Keywords: Critical international normalized ratio Quality indicators Warfarin Communication Oral anticoagulants HAS-BLED score Turnaround time
a b s t r a c t Objective: This study aimed at reviewing the quality indicators for reporting critical international normalized ratio (INR) in a coagulation laboratory. Design and methods: This is a retrospective study conducted at Aga Khan University Hospital, hemostatic laboratory from February 2010 till January 2011. Critical INR was defined as ≥5.0. All critical INRs were rechecked and results were communicated to the doctor or patient. Two quality indicators monitored were % of results communicated to the patient/doctor and % of results that remained critical after re-testing. Results: During the study period, a total of 59,980 INRs were reported. Of these 376 or 0.6% were critical. Successful communication of critical results to the doctor or patient was achieved in 275/376 (73.1%). Overall 353 or 94% (343 initial and 10 re-draw) samples had critical INR on repetition. Twenty five patients of the 240 warfarinized patients with critical INR had mild bleeding. No life threatening bleeding was observed in any patient. Conclusion: We observed poor communication for notifying critical INR results during the study period. Routine repeat analysis of critical INR did not alter results in majority of samples. The study assisted in improving communication in subsequent years. Further work is needed to establish evidence based upper and lower cutoff of critical INR. Effect of replicate analysis on turnaround time and accuracy of results needs evaluation. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Novel oral anticoagulants have provided the clinicians with broader choice for anticoagulation in thrombo-embolic conditions. However, vitamin K antagonists such as warfarin, acenocoumarol and phenprocoumon are still used as the drugs of choice in developing countries because of their cost effectiveness and physician preferences. Warfarin inhibits vitamin K epoxide reductase thereby decreasing the synthesis of vitamin K dependent clotting factors (II, VII, IX, X) and natural anticoagulants (protein C and S). It requires laboratory monitoring which is routinely performed by measuring international normalized ratio (INR). For optimum management, INR is maintained within a narrow therapeutic range of 2–3 [1,2]. Similar to all other anticoagulants, overdosing with
Abbreviations: INR, International normalized ratio; ISO, International Organization for Standardization; JCIA, Joint Commission International Accreditation; WHO, World Health Organization; AKUH, Aga Khan University and Hospital; ILMS, Integrated Laboratory Management System; PT, Prothrombin time; APTT, Activated partial thromboplastin time; TAT, Turnaround time. ⁎ Corresponding author at: Section of Hematology, Department of Pathology and Microbiology, The Aga Khan University and Hospital, Stadium Road, P.O. Box 3500, Karachi 74800, Pakistan. Fax: +92 21 34934294, +92 21 34932095. E-mail addresses: [email protected] (S. Hussain), [email protected] (B. Moiz), [email protected] (B. Afaq).
warfarin can lead to bleeding complications which has an incidence of 10 to 17% per year for all events. Major bleeding is described as fatal or intracranial (documented by imaging), ocular (with blindness), retroperitoneal or intra-articular bleeding [3]. Another criterion is severe bleeding causing a fall of 20 g/L (1.24 mmol/L) hemoglobin or more or requiring ≥2 units of whole/red blood cells [3,4]. The annual incidence of major bleeding is 2 to 5% with 0.5 to 1.0% fatal bleeding [5] and 0.2 to 0.4% intracranial bleeding [6,7]. In 1972, Lundberg was the first to describe the communication of critical laboratory tests results to the physicians for possible intervention [8]. Communicating critical results is usually an established policy for quality practice by many accrediting agencies like the International Organization for Standardization (ISO), the Joint Commission International Accreditation (JCIA) and the World Health Organization (WHO). The Aga Khan University and Hospital (AKUH) is a tertiary care hospital situated in Southern Pakistan. It was accredited by ISO 9001 in June 2000 and by JCIA in September 2006. The system of notification of critical results was established in 2007 in the hospital which was switched to the computerized Integrated Laboratory Management System (ILMS) in 2009. Among other critical results, an INR of five or above was considered reportable to health care providers [9,10]. This cutoff was chosen based on published literature and physicians' experiences with warfarin associated bleeding. It is important for a hemostatic laboratory to report
http://dx.doi.org/10.1016/j.clinbiochem.2014.12.019 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
484
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
high INR so that warfarin therapy can be temporarily discontinued. At the same time, it is imperative that laboratory measurement is repeated to observe if a critical INR becomes non-critical on repetition. This study is aimed at reviewing the performance of coagulation laboratory for the notification of critical INR and agreement of INR measurements on repeat test samples. Secondary objective was to define a cutoff for critical INR results reporting. The review was expected to provide results for further improving clinical services of the laboratory. Materials and methods Setting and critical result reporting The AKUH is a 700 bedded tertiary care academic institute with facilities for large oncology and trauma units as well as a fully functional bone marrow transplant unit. The hemostatic laboratory at AKUH is responsible for testing the blood samples of all admitted patients. Additionally, the laboratory acts as a national referral center and receives specimens through 192 blood collection centers. Approximate daily test volumes are 250 prothrombin time (PT)/INR and 150 activated partial thromboplastin time (APTT) samples. Our lab reports INR with every PT test result by default. Our system of critical reporting was established in 2007 as the hospital was accredited with JCIA. Initially the system did not alert either clinical or laboratory personnel for critical results. In 2009, our information structure was upgraded and the computerized system notified physicians for critical results for admitted patients. However, for outpatients, the laboratory technologist/resident communicated the results to the concerned doctor or patient.
was performed on Sysmex ®cs2000i using Dade® Innovin® reagent with an ISI of 1.02. INR was computed by the instrument by calculating [patient PT / control PT]ISI. A critical INR was defined as 5.0 or above. As per our laboratory policy, all PT and INR results in critical range were repeated. If more than 10% discordance was observed between two results then a fresh sample was requested. Delta check was possible only for in-patients. For out-patients, every lab visit was coded individually and therefore previous test results of the same patient could not be retrieved from the system. All verified critical results for admitted patients were communicated to the health care providers within an hour of analysis and read back policy was followed. Outside referrals were communicated to either doctors or patients themselves. A detailed history of the patient was taken by an on call doctor in the laboratory to evaluate the cause of high INR (see Fig. 2). If the patient was found taking warfarin then he was advised to hold the next dose of warfarin and communicate with his physician on urgent basis. In case of laboratory failure to communicate, each patient was given two additional telephone calls within 24 h. In the absence of a response, the information was entered in the computerized system and the patient was considered unapproachable. Evaluation of performance indicators To review the lab performance, we calculated the percentage of successfully communicated critical results of INR to patients/care givers, the percentage of repeated results that met the criteria for critical results and the percentage of repeated results that significantly drifted from initial test result. These performance indicators were adopted from the College of American Pathologists [11] and were calculated as follows:
Study design This was a retrospective study that was conducted from February 2010 to January 2011. Data collection Overall laboratory process of notification of high INR is detailed in Fig. 1. Briefly, blood samples were collected in 3.2% Na citrate and PT
1. % of successful communication of INR = number of patients who were notified of results / total number of patients with critical INR tests × 100 2. % of critical INR that stays critical = number of critical INR re-tested and stayed critical / total number of critical INR tests × 100 3. % of critical INR that became non-critical = number of critical INR re-tested and changed to non-critical INR / total number of critical INR tests × 100.
Fig. 1. Method of communicating and documenting critical INR as established in 2009.
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
485
Fig. 2. Questionnaire for reporting critical INR (≥5.0).
Evaluation of other variables for patients on warfarin therapy During the study period, various parameters of warfarinized patients were also studied such as indications, adverse effects, duration of treatment and frequency of INR monitoring. Special emphasis was given for the site and severity of bleeding events at the time of reporting critical INR. HAS-BLED score was computed for bleeding patients to assess the risk of bleeding [12]. The scoring was done by summation of one point each for history of hypertension, stroke, liver, renal diseases, labile INR, age N 65 years, antiplatelet agents and alcohol consumption.
ranged from 5 to 15 (mean ± SD; 6.8 ± 1.7) with a male to female ratio of 1:1.4. There were ten pediatric patients with age b 15 years, 140 (37%) were between 15 and 40 years and 226 or 60% patients were above 50 years. Of the 376 critical INR, 240 (63.6%) patients were taking warfarin while eight (2%) had other coagulopathies such as liver dysfunction or vitamin K deficiency. The cause of critical INR remained undetermined in 128 (34.3%) patients due to communication failure (n = 101) or patients' unawareness of their disease/drugs intake (n = 27).
Statistical analysis
Quality indicators for laboratory performance
All data was entered into SPSS version 21.0 (IBM, Armonk, NY, USA). Descriptive data was given as number (%) and frequency for qualitative variables and mean ± standard deviation for quantitative variables. Frequency and percentages were calculated to check the association between bleeding and non-bleeding patients with different parameters, and fisher exact test or chi square test was applied.
Of the 376 critical INR, 275 results were communicated to patients/ health professionals constituting 73% of total critical INR. Reason for communication failure in remaining critical results (n = 101 or 26.9%) was absent or incorrect contact number. As per policy, every abnormal test was repeated on the same sample. Three hundred forty-three (91.2%) results remained unchanged in following retesting. New samples were requested from 33 patients with 12 refusing for redraw. Of the 21 fresh samples obtained, INR changed from a mean critical value of 7.1 ± 0.95 to a non-critical mean of 1.4 ± 0.7 in 11 patients while remaining critical (from an original mean of 7.2 ± 0.94 to 6.6 ± 0.75) in 10 patients. Thus, overall 353 or 94% (343 initial and 10 re-draw) samples had critical INR on repetition. A further analysis showed that 10 of 11 samples that became non critical (b5.0) were received from various blood collecting centers as frozen samples. On investigation, it was found that loss of sample integrity during transport and inappropriate collection technique was responsible for spuriously raised INR in samples received from collection centers. Hence pre-analytic error was the reason behind this high level of disagreement.
Ethical approval The study was approved by the ethical review committee of The Aga Khan University (#2526-Obs-ERC-13). Results Demographics During the study period, 59,980 total and 376 (or 0.6%) critical (≥ 5.0) INRs were reported on outpatient samples. Critical INR
486
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488
Characteristics of warfarinized patients Two hundred and forty of 376 patients with critical INR were taking warfarin. There were 145 females and 95 males with a mean age of 55.7 years ± 18.1 and 89 patients (37%) above the age of 65 years. Seventy four (31%) patients were on lifelong treatment while 77 (32%) and 17 patients were taking warfarin respectively for 0–6 and 6–12 months. Seventy two patients (30%) could not recall the dates of warfarin initiation. Frequency of INR measurement was weekly or biweekly in 73 (30%) and monthly in 49 patients (20%) while no set pattern was observed in the remainder. Valvuloplasty (n = 53 or 22%) was the commonest indication of warfarin intake followed by deep venous thrombosis (n = 47 or 20%) and cerebrovascular accident (n = 42 or 18%). Less common indications included coronary by-pass surgeries/ angioplasty (n = 27 or 11%), pace makers/arrhythmia (n = 17 or 7%), myocardial infarction (n = 11 or 5%) and polycythemia (n = 1) while indications were not known to 42 patients. Majority of the patients (n = 63 or 26%) were on anti-hypertensive medications while 32 or 13% were taking aspirin. Sixty four patients (27%) were taking at least two or more medications beside warfarin at the time of INR testing. Warfarin toxicity and critical INR Of the 240 patients on warfarin, adverse effects were observed in 43 (18%) including bleeding (n = 24), headache (n = 13), diarrhea (n = 5) and combined bleeding and diarrhea (n = 1). Gingival bleeding was the commonest complaint observed in ten patients. Others were bruising (n = 4), hemoptysis (n = 2), hematoma (n = 2), gastrointestinal (n = 1), menorrhagia (n = 4), post-operative (n = 1) and hemarthrosis (n = 1). Majority of hemorrhagic patients (n = 16) were females. Nine patients were on extended warfarin therapy and six and four were taking warfarin secondary to valvuloplasty and CVA respectively. Only three of 25 hemorrhagic patients were taking aspirin. No major or life threatening bleeding was observed. INR was ≥10.0 in three patients and it ranged from 5.0 to ≤ 10.0 in 22 patients [Fig. 3]. Majority (n = 10) of the bleeding patients had INR ranging from 5.0 to ≤ 6.0. HAS-BLED scores were 0, 1, 2 and 3 respectively in 10, 5, 6 and 3 patients with bleeding [Fig. 4]. Table 1 gives comparative analysis of various parameters of two groups of patients with and without bleeding. History of myocardial infarction and INR were statistically significant for bleeding. Characteristics of patients with critical INR due to other causes (n = 136) Though our data focuses on the warfarinized group, information for 136 patients with critical INR due to other causes was also available.
Fig. 4. Comparison of HASBLED score in warfarinized [n = 240] patients with [n = 25] and without bleeding [n = 215].
These were 73 (57%) females and 55 (43%) males. The mean age was 50 ± 17 years and the mean critical INR was 7.1 ± 2.1. No contact was possible in 101/136 cases due to incorrect or no contact number. Of the remaining 35, a large group was ignorant of disease/drug intake (n = 27) while eight patients suffered from coagulopathies. Of those with coagulopathies (n = 8), two patients had bruises, one patient (3 years of age) bled after circumcision and one patient had gastrointestinal bleeding. Hence adverse effects of bleeding were observed in 4/8 or 50% patients emphasizing the significance of communicating critical INR even in non-warfarinized patients. Discussion We observed 0.6% critical INR results during the study period. Successful communication to outside referrals was 73% while 93% test results stayed critical after repeat measurements. We had a high communication failure of 37% for out-patients. Threshold for critical INR In 2007, we identified critical values for various chemistry and hematology tests based on published literature. A critical INR was set at 5.0 or above in our laboratory at which only 10% of the warfarinized patients were minimally hemorrhagic. Their INR ranged from 5 to 15 and HAS-BLED scoring was 0 to 3. No life threatening episodes or major bleeding events were observed in patients communicated to. Most
Fig. 3. Comparison of critical INR in warfarinized [n = 240] patients with [n = 25] and without bleeding [n = 215].
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488 Table 1 Base line characteristics of warfarinized patients [n = 240] with and without bleeding having critical INR. Descriptive values are mean ± 1SD.
487
consequence. According to JCIA, critical results should be informed to clinician who can take immediate action [19].
Variable
Non-bleeding n = 215
Bleeding n = 25
Total n = 240
p-Value
Testing frozen samples and INR reporting
Age N 65 years n (%) Females n (%) Aspirin intake Hypertension Diabetes mellitus Previous stroke Previous myocardial infarction Previous valvuloplasty Previous CABG Atrial fibrillation Monthly INR monitoring Time since warfarin intake in months (mean ± SD) INR (mean ± SD) HAS-BLED score (mean ± SD)
81 (37.7) 128 (59.5) 29 (13.5) 54 (25.1) 12 (5.6) 37 (17) 7 (3.3) 47 (22) 24 (11) 16 (7.4) 118 (54.8) 4.6 ± 2.7
8 (32) 17 (68) 3 (13) 9 (36) 2 (8) 5 (20) 4 (16) 6 (24) 3 (12) 1 (4) 14 (56) 4.0 ± 2.4
89 145 32 63 14 42 11 53 27 17 132 (55) 4.6 ± 2.6
0.774 0.413 0.836 0.242 1.000 0.728 0.015⁎ 0.807 0.728 1.000 0.843 0.409
Frozen plasma samples with elevated INR received from phlebotomy centers could not be differentiated from EDTA, heparinized plasma or even from serum. When coagulation samples are received as frozen ‘plasma’ and show prolonged PT/APTT, CAP [20] recommends testing fibrinogen levels. A very low fibrinogen level below detection limit in absence of warfarin or bleeding history would imply testing on fresh sample. However, due to financial restrain, we do not perform fibrinogen assay on any sample but rather request fresh sample for repeat testing.
6.6 ± 1.2 1.0 ± 1.0
7.1 ± 2.7 1.1 ± 1.0
6.6 ± 1.5 1.0 ± 1.0
0.042⁎ 0.781
⁎ p value less than 0.05 is statistically significant.
laboratories use a critical value threshold of 4 or 5 but a wide range of 2.6 to 10 has been observed [13]. A critical value needs to be optimally chosen and must be evidence based to represent a life threatening situation for reducing mortality [14]. A lower cutoff could not be established by us and patients bleeding on lower INR would not have been captured. Repeating analysis Every critical result was repeated and 91% of our results remained unchanged. It is argued by Niu A and Deetz CO [15,16] that such repetition increases the turnaround time and cost of the testing without any added accuracy. Toll AD in 2011 [17] described differences between two test runs of 553 individual critical PT results. He found a maximum difference of 4 s in 99.4% results. He concluded that repeat testing in critical testing did not provide any advantage in coagulation setting. A recent evaluation of various tests (not including INR) in 86 laboratories showed that routine repeat analysis of automated hematology and chemistry critical values is unlikely to be useful and may adversely affect patient care [18]. Some laboratories rely on delta check to determine the need of repeating. However, in our case this was not possible as our ILMS did not allow tracking of previous test results in outpatients. In our laboratory the mean turnaround time (TAT) for reporting a verified critical result was 45–60 min. At the time, this study was done there was no stress on TAT for reporting. TAT for reporting INR on a fresh sample was 2 h after receiving the sample. Hence where a redraw was required for INR confirmation, reporting would take 24 h after first sample draw. JCIA recommends an explicit time frame of 30 min between completion of test and receipt by responsible care giver. In 2013, ILMS was upgraded to capture this time interval and measured quarterly as a quality indicator of the lab. In 2012, this time line was achieved in 47% of the cases [data not shown] and improved to 97% in 2013. Communication failure This was a major barrier in our situation. Contact numbers provided were either incorrect or patients did not receive calls. Similarly physicians' contact numbers are unknown to patients themselves who visit laboratory for their testing. To handle this situation, it was suggested that patients should be warned at the time of blood draw that they might receive a call from the laboratory in case of critical results. E-mails/fax could be suitable alternative tools of communication. In Pakistan, notification alerts are directed mainly towards patients or his relatives and rarely to care givers. Our experience in communicating results to patients was very difficult as they might not understand the
One critical value for all We had one critical INR value of N 5.0 for all types of patients. In certain situations what is “life-threatening” for one group may not be for all others and vice versa. Working with clinicians is imperative to understand the processes in “unique” locations e.g. an INR of N 1.5 may be critical for a patient with anticipated bleeding such as an invasive procedure [21]. Conclusions We observed poor communication for notifying critical INR results during the study period. Routine repeat analysis of critical INR did not alter results in majority of the samples. The study assisted in improving communication in subsequent years. Further work is needed to establish evidence based upper and lower cutoff of critical INR. Effect of replicate analysis on TAT and accuracy of results needs evaluation. Conflict of interest The authors declare that there are no conflicts of interest. Authors information SH is a fifth year resident in the Hematology Department at the Aga Khan University and Hospital. BM is an associate professor and consultant hematologist at the Aga Khan University and Hospital. BR is the assistant manager in the Coagulation section of the Clinical Laboratory. Author contributions SH analyzed the data and wrote the original draft of the paper. BR analyzed the data and BM provided the original idea for conducting this audit, guided the work, reviewed the paper, analyzed the data and wrote some of the manuscript. All authors were involved in revising and approving the manuscript. Acknowledgments We thank Dr Rabia Rafiq, Mehwish Arshad and Subuhi Raza for collecting the data. We also thank Mr. Ahmad Raheem for analyzing the data. References [1] ICSH/ICTH recommendations for reporting prothrombin time in oral anticoagulant control. International Committee for Standardization in Haematology and International Committee on Thrombosis and Haemostasis. J Clin Pathol Feb 1985;38(2): 133–4 [PubMed PMID: 3968214. Epub 1985/02/01. eng]. [2] Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians
488
[3]
[4]
[5]
[6]
[7]
[8] [9]
[10]
S. Hussain et al. / Clinical Biochemistry 48 (2015) 483–488 Evidence-Based Clinical Practice Guidelines (8th Edition). Chest Jun 2008;133(6 Suppl.):160S–98S [PubMed PMID: 18574265. Epub 2008/07/24. eng]. Schulman S, Kearon C. Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost Apr 2005;3(4): 692–4 [PubMed PMID: 15842354]. Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D'Angelo A, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet Aug 17 1996;348(9025):423–8 [PubMed PMID: 8709780. Epub 1996/08/ 17. eng]. Stehle S, Kirchheiner J, Lazar A, Fuhr U. Pharmacogenetics of oral anticoagulants: a basis for dose individualization. Clin Pharmacokinet 2008;47(9):565–94 [PubMed PMID: 18698879. Epub 2008/08/14. eng]. Schulman S, Beyth RJ, Kearon C, Levine MN. Hemorrhagic complications of anticoagulant and thrombolytic treatment: American College of Chest Physicians EvidenceBased Clinical Practice Guidelines (8th Edition). Chest Jun 2008;133(6 Suppl.): 257S–98S [PubMed PMID: 18574268. Epub 2008/07/24. eng]. Palareti G, Cosmi B. Bleeding with anticoagulation therapy — who is at risk, and how best to identify such patients. Thromb Haemost Aug 2009;102(2):268–78 [PubMed PMID: 19652877. Epub 2009/08/05. eng]. Lundberg GD. When to panic over abnormal values. MLO Med Lab Obs 1972;4: 47–54. Fairweather RB, Ansell J, van den Besselaar AM, Brandt JT, Bussey HI, Poller L, et al. College of American Pathologists Conference XXXI on laboratory monitoring of anticoagulant therapy: laboratory monitoring of oral anticoagulant therapy. Arch Pathol Lab Med Sep 1998;122(9):768–81 [PubMed PMID: 9740135. Epub 1998/09/18. eng]. Hirsh J, Dalen JE, Anderson DR, Poller L, Bussey H, Ansell J, et al. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest Nov 1998;114(5 Suppl.):445S–69S [PubMed PMID: 9822057. Epub 1998/11/20. eng].
[11] 2010_qmt_catalog.pdf. [12] Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest Nov 2010;138(5):1093–100 [PubMed PMID: 20299623]. [13] Lusky K. INR practice gaps found in Q-Probes 2011 [cited 2014 30-10-14]. Available from: http://www.cap.org/apps//cap.portal?_nfpb=true&cntvwrPtlt_ actionOverride=%2Fportlets%2FcontentViewer%2Fshow&_windowLabel= cntvwrPtlt&cntvwrPtlt%7BactionForm.contentReference%7D=cap_today%2F0311% 2F0311c_inr_practice.html&_state=maximized&_pageLabel=cntvwr. [14] Doering TA, Plapp F, Crawford JM. Establishing an evidence base for critical laboratory value thresholds. Am J Clin Pathol Nov 2014;142(5):617–28 [PubMed PMID: 25319976]. [15] Deetz CO, Nolan DK, Scott MG. An examination of the usefulness of repeat testing practices in a large hospital clinical chemistry laboratory. Am J Clin Pathol Jan 2012;137(1):20–5 [PubMed PMID: 22180474]. [16] Niu A, Yan X, Wang L, Min Y, Hu C. Utility and necessity of repeat testing of critical values in the clinical chemistry laboratory. PLoS One 2013;8(11):e80663 [PubMed PMID: 24260448. Pubmed Central PMCID: 3834106]. [17] Toll AD, Liu JM, Gulati G, Behling EM, Kocher WD. Does routine repeat testing of critical values offer any advantage over single testing? Arch Pathol Lab Med Apr 2011;135(4):440–4 [PubMed PMID: 21466359]. [18] Lehman CM, Howanitz PJ, Souers R, Karcher DS. Utility of repeat testing of critical values: a Q-probes analysis of 86 clinical laboratories. Arch Pathol Lab Med Jun 2014;138(6):788–93 [PubMed PMID: 24878017]. [19] Hanna D, Griswold P, Leape LL, Bates DW. Communicating critical test results: safe practice recommendations. Jt Comm J Qual Patient Saf Feb 2005;31(2):68–80 [PubMed PMID: 15791766]. [20] Pathologists CoA. Hematology and coagulation checklistIn: Pathologists CoA, editor. 2011. p. 8. [21] Douketis JD, Bakhsh E. Perioperative management of antithrombotic therapy. Pol Arch Med Wewn Apr 2008;118(4):201–8 [PubMed PMID: 18575419].
