Sie sind auf Seite 1von 5

Optometry (2009) 80, 227-231

Risk factors and complications of subconjunctival

hemorrhages in patients taking warfarin
Lindsey L. Leiker, Pharm.D.,a Bella H. Mehta, Pharm.D.,b Maria C. Pruchnicki, Pharm.D.,b
and Jennifer L. Rodis, Pharm.D.b
a b

Colmery-ONeil Veterans Affairs Medical Center, Topeka, Kansas, and Division of Pharmacy Practice and Administration,
The Ohio State University College of Pharmacy, Columbus, Ohio.

Subconjunctival OBJECTIVES: The aim of this study was to identify patients with subconjunctival hemorrhage (SCH) hemorrhage; on
warfarin therapy, to describe risk factors that may contribute to SCH development, and to identify
Warfarin; complications related to SCH.
Risk factors; METHODS: A retrospective chart review was conducted including patients treated at a university anti-
Complications coagulation clinic over 2 years (4,334 patient visits). Data collection included patient demographics; international
normalized ratios (INRs) before, at time of, and after SCH; risk factors for increased risk of bleeding;
patient-reported complications related to SCH; recent changes in medication use; and warfarin dosage
adjustments made in response to the event. The data were summarized using descriptive statistics and
frequencies described as percentages.
RESULTS: Fifteen SCH events were identified at an event rate of 0.35%. Two were excluded because of
related surgeries near the time of SCH events. The average patient age was 67.3 years (range, 51 to 82).
A total of 76.9% (n 5 10) of patients had INRs within the goal range at the appointment before reporting
the SCH. A total of 46.2% (n 5 6) of patients reported alterations in medication regimens during the
month preceding SCH. Various patient conditions were documented that may increase the risk of SCH
development. No ophthalmic complications were associated with SCHs.
CONCLUSIONS: An SCH event rate of 0.35% was identified. Many factors may have precipitated SCH;
however, ophthalmic complications were uncommon. Optometry 2009;80:227-231

Warfarin sodium is the most widely used anticoagulant variable dose requirements both within and between
in North America for the prevention and treatment of venous individual patients, and the need for strict adherence,
and pulmonary thromboembolisms.1 Despite warfarin is recommended for the prevention and treatment
welldocumented side effects, including bleeding, numerous of multiple thromboembolic disorders.2,3 The effect of
drugfood and drugdrug interactions (including overthe- warfarin is measured via the international normalized ratio
counter medications and herbals), unpredictable and (INR). The INR is a calculated value based on measured
prothrombin time and the individual clinical laboratorys
thromboplastin reagent lot compared with an international
standardized value established by the World Health
Organization.4 Bleeding is listed as the primary adverse effect
Corresponding author: Jennifer L. Rodis, Pharm.D., Division of
Pharmacy Practice and Administration, The Ohio State University College associated with warfarin use and, on an annual basis,
of Pharmacy, Rm. 136e Parks Hall, 500 W. 12th Avenue, Columbus, Ohio. approximately 15% of patients on warfarin will experience

1529-1839/09/$ -see front matter 2009 American Optometric Association. All rights reserved.
228 Optometry, Vol 80, No 5, May 2009
some form of a minor bleeding episode.1 Improved patient documented for patients managed by anticoagulation
outcomes, including decreased occurrence of bleeding and monitoring
greater time within the goal range for INR, is well clinics.5-7
Subconjunctival hemorrhages (SCHs) are among the All patients with documented visits during the specified
minor bleeding-associated adverse effects reported with time period were included. Patients were excluded if they
warfarin use.8-10 Although an SCH may often cause alarm had eye surgery within 3 months of the SCH. (Though eye
because of its disturbing appearance, these anomalies rarely surgery is not a risk factor listed in the literature, it could not
cause symptoms or create a disturbance of the eye and be confirmed whether surgery patients are predisposed to an
usually resolve within 5 to 10 days. Although medical SCH, and thus investigators did not include these patients,
management of the hemorrhage itself is normally not as they were potential confounders.) Study results, described
required, a thorough history is needed to rule out other as frequencies and ranges, are provided for continuous data.
causes. Although trauma is the most common cause, a Although statistical evaluation was planned a priori, because
variety of mechanisms may lead to increased vascular of the low event rate, analysis could not be conducted. This
pressure causing an SCH, including heavy lifting, coughing, chart review was determined exempt by The Ohio State
sneezing, vomiting, or straining. Disease states such as University Biomedical Institutional Review Board.
hypertension, diabetes, bleeding disorders, and Following is a case series of the SCHs identified from this
malignancies may cause an increase in blood volume or review.
viscosity, thereby increasing the risk of SCH
development.9,10 Additionally, acute conditions involving Results
fever, menstruation, acetylsalicylic acid overdose,
anticoagulant induction, and chemotherapy have been cited Of the 4,334 patient visits recorded during the study period,
as causes of SCH.11 Little has been published related to the 15 SCHs were reported resulting in an event rate of 0.35%.
occurrence and complications associated with SCH. SCH was defined as the presence of a bloodshot eye or
Objectives of this study are to identify patients who blood in the eye. Thus, visit notes written in the chart by
presented with SCH on warfarin therapy, describe risk a pharmacist that included this or similar phrasing were
factors that may put a patient at higher risk of SCH identified as SCHs. Eleven patients reported an SCH, 2 of
development, identify the rate of occurrence of whom had more than 1 event. Two of the 15 cases were
complications related to SCH, and determine appropriate excluded, as the patients involved had a history of cataract
health care provider response to this type of event. and related surgeries within a few weeks of the SCH events.
Of the 13 events included in the analysis, the average patient
age was 67.3 years (range, 51 to 82). Individual and group
Methods demographic information is summarized in Tables 1 through
Two pharmacists practicing in a university-affiliated 4. Primary indication for warfarin therapy of patients
ambulatory anticoagulation clinic that manages warfarin included deep venous thrombosis (DVT), pulmonary
therapy as well as anticoagulation bridging with embolism (PE), atrial fibrillation, cerebrovascular accident
lowmolecular-weight heparin performed a retrospective (CVA), and valve replacement.
chart review of patients treated at the clinic during the period Because most patients were unable to come into the clinic
of January 1, 2004, through January 1, 2006. Patients when the event occurred, or did not report the event until
enrolled at the university anticoagulation clinic are seen no their next scheduled appointment, the INR at the time of
less than every 4 weeks for an appointment with a event was not included in the data analysis. A total of 76.9%
pharmacist, which includes a patient interview and (n 5 10) of patient events had an INR within their goal range
assessment, INR measurement, and recommendations at the appointment before reporting the SCH. The 3 patients
related to warfarin dosing and other health-related issues not in range at the anticoagulation clinic appointment before
identified during the appointment. Visit data are the event had supratherapeutic INRs (i.e., INR greater than
documented using an electronic medical record (EMR). individual patient target range). A total of 84.6% of patient
Study data collected from the patient charts included events were within their goal range at the clinic visit
patients age at the time of the event, diagnosis for warfarin immediately after or at the report of the event, with 1 patient
use, INR goal range, most recent INR, current INR (if being supratherapeutic and 1 patient having a subtherapeutic
obtained) at time of event, INR after event, risk factors or INR (i.e., INR below individual patient target range).
comorbid conditions that could increase the overall risk of Medication changes or adjustments, including all
bleeding, and patient-reported complications related to prescription, over-the-counter, and herbal remedies, were
SCH. Recent changes in medication use including new reviewed for up to 1 month preceding the event. A total of
prescription medications, overthe-counter products, dietary 46.2% of patients reported that there was some type of
supplements, and herbs and the specific warfarin dosage alteration in their medication regimen during the month
adjustments made in response to the event were also preceding their event (see Table 1). No complications, such
recorded. as changes in vision, loss of vision, or eye pain or itching,
were reported or associated with any of the SCHs included
in this review. Because of the small event rate identified in were instructed to monitor for these symptoms
this retrospective chart review, statistical associations could andcontactaneyecareproviderifthesesymptomspresented.
not be determined. Although this study was not sufficiently powered to address
Leiker et al Clinical Care 229
Table 1 Individual case demographics

Medicine changes within

Case* Age (yr) Primary diagnosis Secondary diagnosis INR goal range 1 month of event,20

1 51 DVT Protein S deficiency 2-3 d

2 78 Atrial fibrillation d 2-3 d
3 64 CVA Lupus anticoagulants 2-3 d
4 69 DVT Protein S deficiency 2.5-3.5 Started: quetiapine [
Discontinued: escitalopram 4
and trazodone 4
5 53 Mitral valve d 3-4 d replacement
6 69 Atrial fibrillation Mitral valve prolapse 2-3 d
7 79 Atrial fibrillation d 2-3 Amoxicillin x 7 days [
8 56 DVT PE 2-3 d
9 57 DVT PE 2-3 Started: glucosamine [
10 80 PE d 1.5-2.5 Discontinued: procyclidine 4
Decreased dose
of primidone [
11 59 Atrial fibrillation PE 2-3 d
12 82 Atrial fibrillation d 2-3 Discontinued: pantoprazole Y
Started: esomeprazole [
13 69 DVT Protein S deficiency 2.5-3.5 Prednisone burst h21
Started: loratadine 4 and
mometasone 4
Dash depicts nothing reported for this category. Up-pointing arrow indicates this medication may cause an increase in warfarin concentrations and
INR. Horizontal arrow indicates this medication may have no effect on warfarin concentrations and INR. Down-pointing arrow indicates this medication
may cause a decrease in warfarin concentrations and INR. Up-down arrow indicates this medication may cause an increase or a decrease in warfarin
concentrations and INR. DVT, deep venous thrombosis; PE, pulmonary embolism; CVA, cerebral vascular accident. * Eleven patients were identified as
having an SCH with 2 patients reporting 2 SCHs each.
Expected outcome on INR or risk of bleeding with specific medication change.

Discussion statistical associations, it is notable that the patients in cases

4, 7, 9, 10, 12, and 13 started taking medications that could
Currently there is limited literature reporting patient cause an increase in warfarin concentrations and, thus, the
characteristics or complications of SCH related to INR. With an increase in the INR, bleeding complications
concurrent warfarin therapy. Our review over a 2-year could increase as well. However, only the patients in cases
period identified 13episodesofSCH. Nearly 2, 7, and 13 had INRs that were supratherapeutic before the
48%ofpatientshad ahistoryofa comorbid condition that may SCH event. Because warfarin concentrations can be
have increased the risk of SCH, including hypertension, influenced by various lifestyle choices (e.g., acute alcohol
coronary artery disease, breast cancer, historyof stroke, ingestion, diet), it is not possible to determine if the elevated
type2 diabetes, and chronic obstructive pulmonary disease. INRs were related to medications or other potential causes.
None of the events that we reviewed led to ophthalmologic To identify the occurrence of ocular hemorrhage among
complications such as vision loss, double vision, eye pain, warfarin patients, Superstein et al.12 invited anticoagulation
or eye itching. Because no patients reported changes in patients to have their eyes examined in a prospective study.
vision, pain, or itching accompanying the bleeding, patients
Table 3 Individual case results
Five of 126 patients (4.8%) were identified as having some
form of ocular bleeding. Two of those 5 patients had an

Table 2 Group demographics (n 5 11) INR before SCH/ INR after SCH/
Complications Case* in range (yes/no) in range
Average age (yr) 67.3 (51-82) (yes/no) (yes/no)

Primary diagnosis (n, %)

1 2.1/yes 2.2/yes no
DVT 3 (27.3)
2 3.9/no 1.9/yes no
Atrial fibrillation 5 (45.4) 3 2.5/yes 2.0/yes no
4 3.0/yes 1.7/no no
CVA 1 (9.1)
5 4.1/yes 3.4/yes no 6 2.1/yes 2.3/yes
Mitral valve replacement 1 (9.1) no
7 4.4/no 5.2/no no
PE 1 (9.1) 8 3.1/yes 3.0/yes no
9 2.7/yes 2.0/yes no
Secondary diagnosis (n, %) 10 1.6/yes 1.6/yes no
None 5 (45.4) 11 1.8/yes 2.0/yes no
Protein S deficiency 2 (18.2) 12 3.1/yes 2.4/yes no 13 5.4/no 2.4/yes
Lupus anticoagulants 1 (9.1)
* Eleven patients were identified as having an SCH with 2 patients
Mitral valve prolapse 1 (9.1) reporting 2 SCH each.
Complications included changes in vision, loss of vision, eye pain,
or itching.
PE 2 (18.2)

Medication changes within 1 month (n, %)

Yes 6 (46.2) SCH, both of whom were at a therapeutic INR level. These
No 7 (53.8) numbers are comparable with those of 2 older studies from
the 1960s that reported a 1.5% to 5% incidence of SCH
DVT, deep venous thrombosis; PE, pulmonary embolism; CVA,
cerebral vascular accident.
among patients on anticoagulation therapy.13,14

230 Optometry, Vol 80, No 5, May 2009

Several case reports have been published describing

SCHs among patients on (concurrent) warfarin therapy. In
each of these cases, the event was linked to a specific
cause.8,15-19 A recent report of SCH involved an elderly
woman whose SCH could not be connected to any
mechanical causes. No history of recent trauma, illness, or
blood pressure elevations existed, although the patients
INR was supratherapeutic. The INR was eventually
regulated, and the SCH resolved without complications.8
Another report depicts an elderly patient on warfarin therapy
reporting to her optometrist after waking up with eye pain
and what appeared to be a large bruise. Further examination
found normal intraocular pressure, SCH, therapeutic INR, and benign in nature. It is important to understand this and
and blood pressure of 192/90 mmHg, representing an event educate patients accordingly.
that may have been precipitated by elevated blood pressure.
Again, the patient suffered no complications from this
event.15 Acetylsalicylic acid overdose has been linked to the References
SCH of a young man after all other potential causes were
ruled out.16 Bulimia, respiratory distress, and endoscopy 1. Haines ST, Racine E, Zeolla M, et al. Venous thromboembolism. In:
DiPiro JT, Talber RL, Yee GC, et al., eds. Pharmacotherapy: a

Leiker et al Clinical Care 231

have all been linked to case reports of SCH, none of which pathophysiologic approach. 5th ed. New York: The McGraw-Hill
led to ophthalmic complications.17-19 Companies, Inc, 2002:337-73.
2. Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous
Although it appears most SCHs do not lead to ophthalmic thromboembolism: the seventh ACCP conference on antithrombotic
complications, patients should be encouraged to contact a and thrombolytic therapy. Chest 2004;126(suppl):S338-400.
health care professional, such as a primary care or 3. Buller HR, Agnelli G, Hull RD, et al. Antithrombotic therapy for
anticoagulation management provider, in the event of an venousthromboembolicdisease:theseventhACCPconferenceonantithro
mbotic and thrombolytic therapy. Chest 2004;126(suppl):S401-28.
SCH. It is important for health care professionals to attempt
4. Riley RS, Rowe D, Fisher LM. Clinical utilization of the
to identify the cause of the SCH and/or manage secondary internationalnormalized ratio (INR). J Clin Lab Anal 2000;14:101-14.
causes such as uncontrolled hypertension, diabetes, or blood 5. Mehta BH, Rodis JL, Nahata MC, et al. Advancing patient carethrough
dyscrasia. The optometrist may play a key role in identifying innovative practice: the Clinical Partners Program. Am J Health Syst
and managing complications related to an SCH as well as Pharm 2005;62(23):2501-7.
referring patients to other providers for identification and 6. Locke C, Ravnan SL, Patel R, et al. Reduction in warfarin
adverseevents requiring patient hospitalization after implementation of
treatment of secondary causes. Patients on concurrent
a pharmacist-managed anticoagulation service. Pharmacotherapy
warfarin therapy should also have their INR checked.8,11 2005; 25(5):685-9.
7. Ernst ME, Brandt KB. Evaluation of 4 years of clinical pharmacist
anticoagulation ASE management in a rural, private physician office. J
Conclusion Am Pharm Assoc 2003;43(5):630-6.
The risk factors for SCH development include situations that 8. Bodack MI. A warfarin-induced subconjunctival hemorrhage.
Optometry 2007;78:113-8.
may increase venous pressure, not all of which are situations
9. Kabat AG, Sowka JW. Something for nothing. Rev Optom 2006;
limited to patients on anticoagulation therapy. At this time, 143(11):131-2.
supratherapeutic INRs have not been correlated with an 10. Subconjunctival hemorrhage. Rev Optom 2005;142(suppl):A15-6.
increase in the incidence of SCH, although this assumption 11. Wilson RJ. Subconjunctival hemorrhage: overview and management.J
may be made based on the mechanism of injury and known Am Optom Assoc 1986;57(5):376-9.
action of warfarin. Because of the added awareness of 12. Superstein R, Hammouda W, Overbury O. Prevalence of
unusual bleeding that patients on warfarin often experience, ocularhemorrhage in patients receiving warfarin therapy. Can J
Ophthalmol 2000;35:385-9.
patients receiving anticoagulation therapy may become
13. Pollard JW, Hamilton MJ, Christensen NA, et al. Problems
more alarmed should they experience such an event. Patients associatedwith long-term anticoagulant therapy. Circulation
should be adequately educated and assured that the SCH will 1962;25:311-7.
resolve with time, usually without the need for invasive 14. Mosley DH, Schatz IJ, Breneman GM, et al. Long-term
interventions other than INR monitoring and that the risk of anticoagulanttherapy. JAMA 1963;186:128-30.
complications developing as a result is low. Patients should 15. Stead LG, Judson KA. Ocular bleeding due to anticoagulation. N Engl
J Med 2006;355(8):7E.
also be encouraged to contact a health care provider so that
16. Black RA, Bensinger RE. Bilateral subconjunctival hemorrhage
an etiology may be determined and INR checked, especially afteracetylsalicyclic acid overdose. Ann Ophthalmol
if the SCH involves trauma or if pain or vision changes 1982;14(11):1024-5.
occur. Optometrists may note the finding during a routine 17. Weinstein HD, Halabis JA. J Am Optom Assoc 1986;57(5):366-7.
visit, be contacted by patients on warfarin suffering from 18. Chiu CH, Chuang YY, Su LH. Subconjunctival haemorrhage
this condition, or care for patients referred from a primary andrespiratory distress. Lancet 2001;358:724.
19. Rajvanshi P, McDonald GB. Subconjunctival hemorrhage as a
care or anticoagulation provider for ophthalmologic
complication of endoscopy. Gastrointest Endosc 2001;53(2):251-2.
complications. Optometrists should screen for 20. Micromedex Healthcare Series [Internet database]. Greenwood
ophthalmologic complications and refer patients to primary Village, CO: ThomsonHealthcare. Available at:
care or anticoagulation providers for assessment. http://www.thomsonhc. com/home/dispatch. Last accessed March
Our review of the data was limited to a 2-year period, and 26, 2009.
no association between outcomes and measures were 21. Prednisone [database on the Internet]. St. Louis (MO): Drug Facts
andComparisons, eFacts online. c2008 [cited 2008 Apr 25].
apparent. The information provided shows that SCHs do
Available at: Last accessed March 26,
occur in patients on warfarin therapy but are relatively rare 2009.