Photo Credit: Md Babul Hosen
Researchers tested if self-assessments could support the detection of smaller, but still critical, bleeding incidents in patients with immune thrombocytopenia.
Patients with immune thrombocytopenia (ITP) can experience more severe effects from bleeding. While critical bleeds are often captured in bleeding assessments for outpatients with ITP, less severe incidents, like nosebleeds, are more difficult to capture. Patient self-assessments may help capture less severe bleeds; however, questions remain about the accuracy of these assessments.
“Patient-performed bleeding assessments could overcome some of the challenges associated with current methods of bleeding assessments for outpatients with ITP, including completeness and timeliness, but their congruency with assessments done by trained research staff needs to be assessed,” Bianca Clerici and colleagues wrote in the American Journal of Hematology.
To address this need, Clerici and colleagues designed an agreement study comparing bleeding assessments done by trained staff versus by patients. The researchers recruited patients from the McMaster ITP Registry (Hamilton Integrated Research Ethics Board project #15703). This registry includes adult patients with possible thrombocytopenia symptoms who have received care at a hematology clinic.
Clerici and colleagues included 108 patients with primary or secondary thrombocytopenia and a complete bleeding assessment in the study. Most patients (64%) were women, and the median age was 53. Both researchers and patients performed assessments, and the median time between staff and patient assessments was three days.
Bleeding Assessment Results
Trained staff and patients utilized an ITP Bleeding Scale to capture bleeding incidents at nine anatomical areas: nose, skin, mouth, gastrointestinal, genitourinary, gynecological, pulmonary, ocular, and intracranial. Bleeds were most common in the skin (30.2%), epistaxis (28.3%), and mouth (22.9%).
The study team asked patients to grade bleeding as a 0 (none), 1 (mild) or 2 (severe). Since patients weren’t trained, the researchers provided simplified descriptions for each grade. The authors wrote that this description included “possible bleeding signs and symptoms so that these could be classified appropriately.”
The participants ranked 31.5% of bleeds as Grade 1 and 48.1% as Grade 2.
Agreement On Bleeding Assessments
The researchers used two measures to determine the alignment between patient and staff assessments: a 2-way phi agreement and a 3-way weighted kappa agreement. Twenty-two patients (20.4%) experienced no bleeding, while the other 86 (79.4%) had at least one bleeding incident. In patients who experienced bleeding, there were 205 discrete bleeds.
Perfect agreement between staff and patient assessment occurred 40.7% of the time (or for 44 patients). Researchers found excellent congruence for more severe bleeds, including those at the gynecological (k=0.86; CI, 95% 0.71-1.02), gastrointestinal (k=1), genitourinary (k=1), pulmonary (k=1) and intracranial (k=1) sites.
Variability between patient and staff assessments was more likely for less severe bleeds and those that occurred at more common sites. The two-way agreement was good for skin (k=0.68; CI, 95% 0.54-0.82), oral (k=0.76; CI, 95% 0.53-0.98), and ocular (k=0.66; CI, 95% 0.04-1-28) bleeds; and moderate for epistaxis (k=0.58; CI, 95% 0.21-0.95) (Table).
Regarding the relative lack of congruence for less severe bleeds, Clerici and colleagues wrote, “This may reflect a lesser concern for minor bleeds among patients and a greater desire for comprehensive documentation among research staff.”
They also suggested that nose, skin, and epistaxis bleeds may require additional prompts or descriptions for more accurate self-assessments.
Implications For Research & Clinical Practice
Clerici and colleagues concluded that patient self-assessments may be useful in research and clinical settings.
Despite these promising findings, the study had some limitations. Conditions could have changed in the time delay between the staff and patient assessments; the median delay was three days, but the maximum time difference was 13 days. In addition, the results were from a tertiary hematology clinic, so the findings may not apply to other settings.
“Patient self-assessments could enhance comprehensiveness, patient engagement, and real-time data collection for research purposes,” Clerici and colleagues concluded. “With support, this strategy of bleeding self-assessments by patients could also be used in routine clinical practice, which would simplify patient visits, allow for the relay of real-time information, and help identify patients who require urgent attention.”
