Preferences and Values for Treatment Goals in Patients with Acute Ischemic Stroke: Medical Staff Perspectives in China

Background: Individual values and beliefs vary according to cultural and social factors. To better understand clinical decision-making and inform future research, we undertook a survey of preferences in treatment goals for managing patients with acute ischemic stroke among medical professionals in China. Methods: We designed a web-based survey through literature search, face-to-face interviews, pilot questionnaire development, and focus group meetings. Key outcomes explored were death, disability, quality of life, neurological impairments, cognitive dysfunction, and hemorrhagic transformation. Participants were asked to rate the importance of each outcome on a 5-point Likert scale to allow calculation of a weighted score of importance: higher scores indicating greater importance. Results: During promotion among 2700 delegates at several major neurology conferences in 2018, 1189 participants (mean age 40 years, 53% female) completed the survey of whom 96% were clinicians, mainly from tertiary care hospitals located in 30 Provinces of China. For established therapies with bleeding risk (e.g. thrombolysis and thrombectomy), death was the most important outcome (weighted score 4.60), followed by disability (4.45), quality of life (4.45), neurological impairments (4.34), cognitive dysfunction (4.03), and hemorrhagic transformation (3.99). For other acute therapies without bleeding risk, quality of life ranked rst (4.08), followed by disability (4.03), neurological impairment (3.92), death (3.79), cognitive dysfunction (3.78), and hemorrhagic transformation (3.51). Given a therapy with insucient evidence of benet, 845 (71%) participants thought it would be worth trying for potential to improve neurological recovery or quality of life. Conclusions: From the perspective of Chinese medical staff, death is the most important outcome measure for patients with acute ischemic stroke. When reduction of death or disability were less likely to be achievable, the improvement in quality of life is an acceptable outcome measure. In contrast to the previously perceived high fear of bleeding in practice, hemorrhagic transformation was the least concerned even for therapies with a potential bleeding risk.


Introduction
While death and disability are the primary outcome measures used to assess the effects of therapies for acute ischemic stroke (AIS) in randomized controlled trials (RTCs) [1][2][3], the action of clinicians in interpreting such results depends upon many factors, such as the size, consistency and scope of net bene t over risks, and affordability of treatment. Guidelines recommend thrombolysis and thrombectomy for AIS based on a large body of evidence from RCTs [4,5], However, these therapies are only available to a small proportion (< 20%) of AIS patients worldwide, and even fewer (< 10%) in China and other parts of Asia [6][7][8][9][10][11], where neuroprotective agents, herbal products, and other therapies are more widely available and popular [6,12,13] despite insu cient supporting evidence. Reasons for the discrepancy between scienti c evidence and clinical practice are complex.
The choice of treatment therapies is a joint decision between physicians and nurses, and patients and/or their families, where values and preferences over the importance of different outcomes from AIS vary across cultures, experiences and roles. The importance of patient-centered outcomes is now well accepted [14,15], but few studies have been undertaken on the topic in China. The aim of our study was to determine the preferences and values of Chinese medical staff place upon key clinical outcomes used across treatments for AIS.

Methods
We conducted a staged study (Fig. 1 Table S1) for testing among 71 neurologists from 17 hospitals. In a round-table discussion, they had to choose 10 key outcome measures and provide feedback on design of the questionnaire.

Stage 3: Focus group meetings
A focus group of 4 stroke neurologists (ML, BW, SZ, and SW) nalized the survey content based upon the clinical importance of each outcome measure in sections: (i) baseline information; (ii) ranking importance of outcome measures in relation to established acute therapies with bleeding risk (thrombolysis, thrombectomy and antithrombotic agents) and other common therapies of neuroprotective agents, rehabilitation, acupuncture, Chinese patent medicine, and traditional Chinese medicine; (iii) three questions on "If an acute therapy may not reduce death or disability at 3 months but improves neurological de cits in short-term, do you think it is worth trying for stroke patients during the acute phase?" and if 'Yes', "which therapies would you like to use?" and "Please specify the outcome measures that you expect to improve with the chosen therapies".

Stage 4: Web-based surveys
An online questionnaire developed by SurveyStar (Changsha, China) was delivered at 1 national and 6 regional academic, neurology conferences in China from April 2018 to September 2018. A two-dimensional code linked to the questionnaire was promoted in lecture slide presentations and on postcards at the conference venues. Delegates were invited to scan the code and complete the questionnaire via mobile phones.

Statistical analysis
Two researchers (YW, SW) independently coded all interview data and performed thematic analysis of the pilot questionnaire [16]. The web questionnaire was then nalized after comparisons of the included categories with disagreements agreed through consensus. Quantitative data are presented as means and standard deviations for continuous variables, and percentage frequencies for categorical variables. To quantify the importance of each outcome measure ranked by participants, a 5-point Likert scale [17] was used: not important at all = 1, not important = 2, neutral = 3, important = 4, very important = 5. A weighted score was calculated using the following equation: (1*a + 2*b + 3*c + 4*d + 5*e) / (a + b + c + d + e), where a, b, c, d, and e represent participants' numbers for ranked each outcome measure. A higher weighted score equated to greater clinical importance. All statistical analyses were performed in SPSS 25.0 (IBM, Chicago, IL, USA).
There were 845 (71%) participants who answered 'Yes' to the question on treatments with potential to improve neurological impairment but without a bene t on death or disability, with neuroprotective agents (57%), acupuncture (39%), Chinese patent medicine (24%), and traditional Chinese medicine (12%) being the most commonly chosen therapies. The key outcome measure targeted in the use of these therapies were limb function (77%), medical complication (de ned as pneumonia, deep venous thrombosis or urinary tract infection) (65%), level of consciousness (64%), and pathological changes on brain imaging (de ned as infarction volume expansion, reduction of hemorrhagic transformation, brain edema or middle shift) (56%) (Fig. 3).

Discussion
In our study of the preferences and values of health professionals involved in the management of AIS in China, we have shown that death is the outcome of most importance in the use of reperfusion therapies which are proven to reduce death and disability, whilst quality of life is also important. Hemorrhagic transformation had a low priority for these therapies which have a bleeding risk.
The nding that death had a higher priority than disability supports the old adage in Asian cultures that 'better to live than to die', whilst the ability to maintain functional independence in everyday life seems more important in Western culture. For example, as a life-saving intervention for malignant cerebral infarction, decompressive craniectomy was only acceptable to 7% of the general population in Germany, because the potential downside was the chance to continue living with severe disability [18]. However, recognition that people can change their views when affected is re ected in other studies in other countries showing higher rates of acceptance of decompressive hemicraniectomy: 28% in AIS patients, 47% in relatives [19], 16% in nurses [20], and 39% in physicians [21]. In regard to low-risk therapies, neuroprotection, rehabilitation, acupuncture, Chinese patent medicine, and traditional Chinese medicine, are all widely used alongside Western medicines in China [6], despite their limited evidence from RCTs. Our study shows these therapies are primarily used with the intent to improve quality of life.
Due to concerns over treatment-related hemorrhagic transformation, many Asian clinicians prefer to use lower doses of intravenous alteplase for thrombolysis treatment of AIS patients[22-24]. However, our study suggests this adverse event is regarded as less important than the other clinical outcome measures, which is consistent with ndings of dichotomized positive outcomes outweighing the risks of these treatments in RCTs. Education of doctors and patients over hemorrhagic transformation being an acceptable complication in the context of the poor prognosis of AIS may improve the update of reperfusion therapies.
Our web-based survey was able to gather opinions from a large number of health professionals at relatively low cost but is limited by selection bias from including those who were highly skilled and experienced, mainly clinicians, who had the opportunity to attend certain neurology conferences.

Conclusions
In summary, our survey of Chinese medical staff shows they rank death as the outcome of most importance for the use of proven reperfusion therapies, and contrary to current perceptions, the risk of hemorrhagic transformation had a low priority. When a therapy is unlikely to reduce the chances of death or disability, quality of life is more meaningful and acceptable outcome to target. These results may help improve the design of stroke RCTs and counselling of patients and families.