Perceived Impact of Extreme Temperatures, Severe Weather and Air Pollution on Health and Daily Life Among Patients with Cardiovascular Disease: Protocol for a Cross-Sectional Questionnaire-Based Study

Abstract

Background: Extreme temperatures, severe weather events, and air pollution pose important challenges to cardiovascular health. However, patients’ perceptions of environmental risks, symptom-related experiences, adaptive behaviours, and information needs remain insufficiently characterized. This study aims to develop and describe the methodology of a patient-centred questionnaire assessing these dimensions in adults with cardiovascular disease. Methods: An anonymous, self-administered, 37-item questionnaire was developed for adult patients with at least one cardiovascular diagnosis. The instrument comprises five sections addressing sociodemographic and clinical characteristics; perceived health effects of heat, cold, and air pollution; psychological and quality-of-life impacts; protective behaviours and healthcare utilization; and knowledge and information sources. Most closed-ended items use five-point Likert-type or frequency scales, alongside one open-ended question. Content validity was assessed by an independent panel of three experts in cardiology, psychology, and meteorology/climatology. A pilot study involving 39 patients with cardiovascular disease recruited from a clinical research database was conducted to assess questionnaire feasibility and internal consistency. Pilot participants were excluded from the main study. The target sample size for the main study was calculated as 385 participants, increased to 428 to account for non-submission. Eligible patients will be recruited consecutively during hospitalization in our cardiology department. Conclusions: This questionnaire provides a structured framework for investigating perceived environmental cardiovascular vulnerability, psychological impact, adaptive responses, healthcare-related consequences, and information needs. Its application may help identify gaps in patient awareness and protective practices and inform the development of targeted educational and preventive interventions addressing climate-related cardiovascular risks.

1. Introduction

Europe is warming at twice the global average rate [1]. Between the periods 1991–2000 and 2015–2024, heat-related mortality increased in 99.6% of the European regions monitored, by a mean of 52 annual deaths per million inhabitants (95% CI, 43–59). Heat-related deaths in 2024 were estimated at 62,775 [2]. Over the same period, daily health warnings for extreme heat increased by 318% across Europe and by 198% in eastern Europe, where older adults experienced the highest heatwave exposure in 2024, averaging 34.5 heatwave days per person [2]. Older adults and people with pre-existing chronic conditions are among the groups at greatest risk [1].

Romania, particularly its capital, is also strongly affected. A rapid attribution analysis of 854 European cities estimated 24,404 excess heat-related deaths during the summer of 2025, of which 68% were attributable to human-induced climate change [3]. Among the 35 Romanian cities included in the analysis, 1487 excess deaths were estimated, including 472 in Bucharest. The latter corresponds to an age-standardised excess mortality rate of 352 per million inhabitants, the third-highest among 30 European capitals, after Rome and Athens [3]. Between the periods 1991–2000 and 2015–2024, heat-related mortality in eastern European cities increased by 98 annual deaths per million inhabitants (95% CI, 54–140) [4].

Patients with cardiovascular disease are particularly vulnerable to these environmental exposures. Extreme heat can trigger acute myocardial and cerebrovascular events and exacerbate chronic conditions [5]. Cold exposure is also relevant: in a nationwide Swedish case-crossover study involving 482,000 heart failure hospitalisations, cold spells were associated with higher odds of hospitalisation over the subsequent 2–6 days (OR, 1.085; 95% CI, 1.042–1.129), while higher warm-season temperatures were associated with higher odds over the subsequent 0–3 days [6]. Air pollution poses an additional cardiovascular threat. Fine particulate matter (PM_2.5) is a recognised risk factor for premature mortality and respiratory and cardiovascular disease [1], and air pollution has been associated with hypertension, atherosclerosis, acute myocardial infarction and stroke [7]. Environmental pollution has also been examined in relation to peripheral artery disease (PAD): an ecological cross-country analysis suggested a correlation between PM₁₀ concentrations and PAD prevalence, a finding presented by its authors as hypothesis-generating [8]. In eastern European cities, where solid fuels are widely used for residential heating, the PM_2.5-attributable mortality rate in 2023 was more than three times that in northern European cities (153 vs. 45 per 100,000 inhabitants) [4]. These exposures may also interact: heat and sunlight promote ground-level ozone formation, while wildfires and dust storms occurring during heatwaves can increase particulate matter concentrations [9].

Despite the established health risks, less is known about how patients themselves perceive these exposures and respond to them. Published surveys have primarily focused on the general population [5,10,11,12,13] or on healthcare professionals and students [7,14], and have generally examined climate change and heat [5,13,14,15] or air pollution [7,10,11,12] separately. Few studies have investigated patients’ perspectives on the health effects of climate change [15], and the need for research targeting vulnerable groups, including people with cardiovascular disease, has been explicitly recognised [12]. To our knowledge, none of the instruments reviewed for the present study was specifically designed for patients with cardiovascular disease or addressed, within a single questionnaire, the perceived effects of heat, cold and air pollution on symptoms alongside their emotional impact, protective behaviours, healthcare utilisation and information needs.

The present study aims to address this gap through a cross-sectional, questionnaire-based investigation of adult patients with cardiovascular disease. Its primary objective is to describe patients’ perceptions of the effects of extreme temperatures (heat and cold), severe weather events and air pollution on their symptoms, general health and daily lives. Secondary objectives are to describe (i) the psychological and emotional impact of these exposures; (ii) protective behaviours, self-reported treatment adjustments and healthcare utilisation; (iii) patients’ self-rated knowledge and information needs, sources of information, and perceptions of the treating physician’s role; and (iv) associations between these perceptions and sociodemographic and health-related characteristics.

2. Materials and Methods

2.1. Study Design and Status

This is an observational, cross-sectional, questionnaire-based study. The questionnaire has been pretested in a pilot sample (Section 2.9). Enrolment of the main study sample began in October 2026 and is ongoing.

2.2. Setting

This single-centre study is conducted in the Department of Cardiology of Clinical Emergency Hospital of Bucharest (SCUB), Bucharest, Romania. Enrolment has no fixed end date and will continue until the target sample size is reached (Section 2.5).

2.3. Participants

The study is conducted in accordance with the ethical principles of the Declaration of Helsinki. Participation is voluntary and anonymous, and informed consent is obtained from all participants before enrollment. The study protocol received approval from the SCUB Ethics Committee (decision no. 53916/2026, 23 September 2026).

The target population comprises adult patients with cardiovascular disease. Patients are eligible if they are aged 18 years or older, are admitted to the Department of Cardiology during the enrolment period, and have at least one cardiovascular diagnosis documented in their medical record. A member of the study team verifies the diagnosis in the medical record before inviting the patient to participate.

No specific cardiovascular diagnosis is required for inclusion, and no particular cardiovascular diagnosis constitutes an exclusion criterion. Cardiovascular diseases are defined broadly according to Chapter IX of the International Classification of Diseases, 10th Revision (ICD-10), which covers diseases of the circulatory system (codes I00–I99) [16]. This broad definition reflects the descriptive aim of characterising perceptions across the patients of a general cardiology department. The included conditions differ in pathophysiology and probably in environmental vulnerability (for example, heart failure, coronary artery disease, arrhythmias, hypertension and peripheral artery disease); the implications of this heterogeneity are addressed in Section 2.12 and Section 4.1.

Patients are excluded if they (i) decline to participate; (ii) have insufficient Romanian language proficiency to complete the questionnaire; (iii) have a neurocognitive disorder that prevents them from understanding the questions; (iv) are clinically unstable at the time of recruitment, in which case they may be approached again once their condition has stabilised; or (v) have already completed the questionnaire, either during the pilot phase or during an earlier admission within the enrolment period. Inability to use a mobile phone is not an exclusion criterion, as assisted questionnaire completion is available (Section 2.10).

2.4. Recruitment

All eligible patients admitted to the department are invited to participate consecutively. Before extending the invitation, a member of the study team verifies eligibility against the inclusion and exclusion criteria, including confirmation of the cardiovascular diagnosis in the medical record. Patients may be approached at any point during their hospital stay. To allow a recruitment flow diagram to be reported (Section 4.2), the study team keeps an anonymous screening log that records, without any identifiers, the numbers of patients screened, eligible and invited, the reasons for ineligibility, the number who declined, and the number of submitted questionnaires.

2.5. Sample Size

The required sample size was calculated using the formula for estimating a population proportion: n = Z² × p × (1 − p)/d², where (Z = 1.96) represents the standard normal deviate for a 95% confidence level, (p) is the anticipated proportion, and (d) is the desired absolute precision [17]. In the absence of a reliable prior estimate of the proportion of interest among patients with cardiovascular disease, (p) was set at 0.50, which maximizes the required sample size. The absolute precision was set at 0.05. The resulting minimum sample size was 384.16; rounding up yielded a target of 385 completed questionnaires. This calculation ensures that any proportion estimated in the full sample (for example, the proportion of participants who agree that heat worsens their cardiac symptoms) will have a 95% confidence interval no wider than ±5 percentage points. The study is not powered for subgroup comparisons or for detecting associations of a prespecified size: in subgroups of 100 and 50 participants, the corresponding precision would be approximately ±9.8 and ±13.9 percentage points, respectively. Subgroup and multivariable analyses (Section 2.12) are therefore exploratory, and their results will be interpreted with caution.

All items in the electronic form are mandatory, so every submitted questionnaire is complete (Section 2.10), and item-level missing data are not possible. To allow for an assumed 10% of enrolled patients who consent but do not submit the questionnaire (for example, because of discharge, fatigue or clinical deterioration before completion), the enrolment target was set at 428 patients (385/0.90 = 427.8). No finite-population correction was applied, because enrolment is not limited to a fixed period.

2.6. Questionnaire Development

The questionnaire was developed in Romanian by three experts: a cardiologist, a clinical psychologist and a physicist with expertise in environmental science. Item development was based on a review of published survey instruments on perceptions of climate change, extreme weather and air pollution [5,7,10,11,12,13,14,15], summarised in Table 1.

Table 1. Source instruments reviewed during questionnaire development.

Ref.	Instrument and Country	Respondents and Mode	Content Relevant to the Present Questionnaire
[5]	TK Monitor of Patient Safety 2023, Germany	1000 adults, general population; telephone interview	Concern about climate change affecting one’s own health; perceived need for medical treatment because of hot weather; effect of heat waves on chronic illness; wish for counselling by the general practitioner
[7]	Knowledge, attitudes and practices regarding air pollution, Colombia	352 medical students; self-administered online form	Perceived air quality; knowledge of health effects, including cardiovascular disease; interest in training; perceived ineffectiveness of one’s own actions; protective practices
[10]	Knowledge, attitudes and perception of air pollution, Ireland	1005 adults, general population; telephone	Awareness of air pollution and its health effects; sociodemographic determinants
[11]	Public knowledge of air pollution, Poland	1092 adults, general population; web interview	Health consequences and groups at risk; marital status and children as covariates
[12]	Perceptions and awareness of air quality, Thessaloniki, Greece	245 adults; online questionnaire	Perceived air quality; awareness; stress and anxiety; impact on outdoor exercise; smoking; time spent outdoors; respiratory or heart disease; physician advice
[13]	Survey of environmental concerns, part of a doctoral thesis, United Kingdom	General public; postal questionnaire	Whether air pollution has affected one’s own health; sources of information; five-point agree–disagree statements, including personal efficacy
[14]	Knowledge, attitudes and practices on climate change, India	3062 members of the healthcare workforce; telephone, online and in person	Heat-related and cold-related illness; sources of information; interest in learning more
[15]	“Sustainability at the Hospital” survey, Germany	399 hospital patients; online survey accessed by QR code	Patient-level sociodemographic items (relationship status, children); five-point agreement scales in a patient sample

2.7. Questionnaire Structure

The questionnaire begins with a brief introduction outlining the study’s purpose and emphasizing that there are no right or wrong answers and that participation is voluntary and anonymous. The estimated completion time is 10–12 min. The instrument comprises 37 items organized into five sections (Table 2). Items 12–28 and 30–36 use five-point response scales, and each item is displayed in the form with its own fully labelled response options. Four sets of anchors are used: agreement (1 = strongly disagree to 5 = strongly agree; items 12–25 and 35); frequency (1 = never to 5 = always; items 26–28 and 34); self-rated information (1 = not at all informed to 5 = very well informed; items 30 and 31); and intensity (1 = not at all to 5 = very much; items 32, 33 and 36). Item 29 uses five ordered categories, ranging from no change made or intended to major changes made (Table 2 and Supplementary File S1).

Table 2. Overview of questionnaire sections, items and response formats.

Section	Items	Content	Response Format
1. Sociodemographic and health data	1–11	Age group, sex, urban or rural residence, education, current occupation, marital status, children, diagnosed chronic conditions, smoking status, physical activity, daily time spent outdoors	Closed categories; occupation as free text; chronic conditions as multiple response
2. Perceived impact of extreme temperatures and air pollution on health	12–18	Perceived local air quality; worsening of cardiac symptoms by heat and by intense cold; effect of air pollution on general health; worry; perceived link between pollution episodes and symptoms; perceived personal sensitivity	Five-point agreement scale
3. Psychological and emotional impact	19–22	Stress or anxiety with sudden weather changes; helplessness regarding pollution; effect on quality of life; perceived control over protecting one’s health	Five-point agreement scale
4. Attitudes, behaviours, use of health services and life changes	23–29	Avoiding going outdoors; adapting the daily routine; following weather and air-quality warnings (23–25); treatment adjustment, unscheduled consultations, hospital admission or emergency assessment (26–28); major life changes made or intended (29)	Five-point agreement scale (23–25); five-point frequency scale (26–28); five ordered categories (29)
5. Knowledge and sources of information	30–37	Self-rated information on extreme weather and cardiovascular health and on pollution and health; concern about environmental factors; information received from the physician; use of online sources; usefulness of mass media; interest in educational programmes; one useful source named by the respondent (37)	Five-point scales with item-specific anchors; information level (30–31), intensity (32, 33, 36), frequency (34), agreement (35); free text (37)

Several design features should be noted. Age (item 1) is recorded in predefined groups because the study aims to compare perceptions across age groups rather than to model age as a continuous predictor; the upper cut-off (≥75 years) matches the age stratification used in large European studies of heat-related mortality, in which risk was highest in the oldest age groups [18,19]. Item 8 is intended to capture comorbidities, particularly respiratory and metabolic diseases, which may modify environmental vulnerability; its cardiovascular option completes the list of chronic conditions and is not used to establish eligibility, which is verified in the medical record. Items 13–15 and 17 address heat, cold and air pollution separately, whereas items 16, 18, 21, 23–28 and 32–35 refer jointly to extreme weather and/or air pollution, so as to capture the overall perceived burden of environmental stressors without lengthening the questionnaire; these items will be interpreted as such, not as exposure-specific measures. Items 26–28 do not specify a recall period and therefore describe the respondent’s overall experience over the course of the disease, and item 26 does not distinguish treatment changes directed by a physician from those made by patients themselves. Item 29 is a single ordered scale, from no intention to act, through consideration, to action of increasing extent, broadly informed by the transtheoretical model [20]. No “don’t know” or “not applicable” option is provided, so respondents who cannot judge an effect may have selected the neutral midpoint of the agreement scales. These features were retained because the questionnaire had been reviewed by the expert panel, pretested and approved by the Ethics Committee in its present form, and enrolment with this version had begun; their implications for analysis and interpretation are addressed in Section 2.12 and Section 4.1

2.8. Content Validity

Content validity was assessed by an independent panel of three experts, covering the main components of the construct—a cardiologist, a psychologist, and an expert in meteorology and climatology—none of whom had been involved in drafting the questionnaire items. Each expert evaluated every item individually on a structured evaluation form and rated it for relevance and for clarity on four-point scales:

• Relevance: 1 = not relevant; 2 = somewhat relevant (major revision needed); 3 = relevant (minor revision needed); 4 = highly relevant;
• Clarity: 1 = not clear; 2 = somewhat clear; 3 = clear (minor revision needed); 4 = very clear.

The experts could also suggest rewording, removing or adding items, and gave an overall assessment of whether the five-section structure and the length of the questionnaire (37 items) were appropriate. All 37 items achieved an I-CVI of 1.00 (ratings of 3–4 considered acceptable) for both relevance and clarity (S-CVI/Ave = 1.00); only one relevance rating of 3 was given, without suggestions for revision, and all clarity ratings were 4. No changes to the questionnaire were required after expert review. With three experts, the only possible I-CVI values are 0, 0.33, 0.67 and 1.00, and, according to Lynn’s criteria discussed by Polit, Beck and Owen [21], an item is acceptable only if all experts rate it as relevant (I-CVI = 1.00) when the panel has five or fewer members. A value of 1.00 is therefore the minimum acceptable value for this panel rather than an exceptional result. The indices reflect agreement within a small panel and constitute preliminary evidence of content validity, not comprehensive validation; construct validity has not yet been established and will be examined in the main study (Section 2.12).

2.9. Pilot Testing and Internal Consistency

The questionnaire was pretested in a pilot sample of 39 participants. Patients with cardiovascular diseases who had previously expressed willingness to participate in research conducted at the clinic were invited to participate in the pilot study. Pilot participants are not eligible for the main study (exclusion criterion v).

Internal consistency was assessed with Cronbach’s alpha, calculated separately for each thematic section of the questionnaire (questionnaire Section 2, Section 3, Section 4 and Section 5; Table 2) as a preliminary check. Cronbach’s alpha was 0.82 for questionnaire Section 2 (items 12–18), 0.70 for questionnaire Section 3 (items 19–22), 0.75 for questionnaire Section 4 (items 23–29) and 0.76 for questionnaire Section 5 (items 30–36), indicating acceptable-to-good internal consistency. The sections were defined a priori as thematic content domains, based on the literature review and the development work of the expert team, rather than as unidimensional scales measuring a single latent construct. Given the small pilot sample, which did not allow factor analysis, these coefficients provide only preliminary evidence of internal consistency and do not establish the dimensional structure or construct validity of the instrument. The questionnaire was retained in its pretested form for the main study.

2.10. Data Collection

The questionnaire is administered electronically using Google Forms. After providing informed consent, patients access the form by scanning a QR code with a mobile phone and complete it independently. Patients who do not own a smartphone or are unable to use one complete the questionnaire on a study-provided phone, with assistance from a member of the study team. This alternative was included to reduce the risk of excluding older adults and individuals with lower educational attainment. The mode of completion (self-administered or assisted) is not recorded in the questionnaire. The form does not collect email addresses or require respondents to sign in.

The questionnaire is anonymous and contains no fields for names, personal identification numbers, addresses or contact details. No clinical data are extracted from medical records. Medical records are consulted solely to verify eligibility before patients are invited to participate; all health-related data analysed in the study are self-reported by participants through the questionnaire. Diagnoses and comorbidities are not abstracted from medical records because questionnaire responses are anonymous and cannot be linked to individual records; such linkage would require the collection of identifiers, which the anonymous design approved by the Ethics Committee was intended to avoid and which might make patients less willing to report their perceptions openly. Self-report is also consistent with the study’s focus on perceptions, as self-reported comorbidities reflect patients’ own understanding of their health.

2.11. Outcomes and Variables

The outcomes are the responses to items 12–36, described by section: perceived health impact (items 12–18), psychological and emotional impact (items 19–22), behaviors, use of health services and life changes (items 23–29), and information level and sources (items 30–36). Answers to the open-ended item 37 will be reported descriptively. The explanatory variables are the sociodemographic and health characteristics recorded in items 1–11. Responses to item 5 will be coded by two investigators into occupational categories (predominantly outdoor work, predominantly indoor work, retired, and other), with answers that cannot be classified assigned to “other”. The proportion of respondents who do not select the cardiovascular option of item 8 will be reported descriptively as a possible indicator of patients’ awareness of their diagnosis. As the study is descriptive and exploratory, no primary outcome was prespecified; the main exploratory analysis will examine the relationship between perceived vulnerability (items 13–15, 17 and 18) and reported protective behaviours (items 23–25).

The study measures perceived exposure and perceived symptom aggravation, not actual exposure. Perception is approached as a psychological construct rather than as a proxy for actual exposure, because perceived threat, rather than objective exposure alone, guides coping and care-seeking behaviour [22,23]. Items 13, 14 and 17 therefore measure patients’ attribution of symptoms to heat, cold or air pollution, not a verified temporal or causal relationship. This attribution is relevant even when it is inaccurate, because it is likely to guide protective behaviour and care-seeking (items 23–28).

2.12. Statistical Analysis

Analyses will be conducted primarily at the level of individual items. Categorical variables will be summarised as frequencies and percentages. Responses to five-point scales will be described using response distributions and, where appropriate, medians and interquartile ranges. For items 26–28, which have no defined recall period, the main analysis will compare respondents who report never experiencing the event with those who report experiencing it at least rarely, and the full response distribution will also be reported. Item 29 will be analysed separately as an ordered categorical variable. Because all items are mandatory, item-level missing data are not possible. Uncodable free-text answers to item 37 will be reported as a separate category. The “prefer not to answer” option for sex (item 2) will be reported as a separate category, and these respondents will be excluded from models that include sex.

Responses to the open-ended item 37 will be analysed using content analysis. Two investigators will independently code each response into categories of information source, starting from an initial framework (healthcare professionals; television and radio; printed press; websites and online news; social media; official meteorological or air-quality services; family and friends; other) and adding categories inductively where needed; responses indicating that no source can be named will be coded separately. Disagreements will be resolved by discussion, and the frequency of each category will be reported.

Associations between sociodemographic or health-related characteristics and questionnaire responses will be explored using the chi-square test or Fisher’s exact test, as appropriate, for categorical variables. Comparisons of ordinal responses between participant groups will use the Mann–Whitney U test or the Kruskal–Wallis test, depending on the number of groups, and correlations between ordinal variables will be assessed using Spearman’s rank correlation coefficient. The analyses will be exploratory and will focus on differences in perceptions, emotional impact, protective behaviours, healthcare utilisation and information needs across participant subgroups.

To account for potential confounding, factors associated with perceived environmental vulnerability (items 13–15, 17 and 18) and with protective behaviours (items 23–25) will be examined using multivariable ordinal (proportional-odds) logistic regression, with the item response as the dependent variable. The explanatory variables, selected a priori, are age group (treated as an ordinal variable), sex, area of residence, education, occupational category, respiratory and metabolic comorbidities, smoking status, physical activity and daily time spent outdoors; sparse categories will be merged. In models of protective behaviours, perceived vulnerability will also be included, as the score derived from the factor analysis if supported, and otherwise as items 13–15, 17 and 18. The proportional-odds assumption will be checked using the Brant test [24]; if it is violated, responses will be dichotomised (agree or strongly agree vs. all other responses) and analysed using binary logistic regression. Results will be reported as odds ratios with 95% confidence intervals. Two interactions of a priori interest (age group × sex and age group × respiratory comorbidity) will be explored. Because specific cardiovascular diagnoses are not recorded, disease-specific subgroup analyses are not feasible.

Because many items and participant characteristics will be examined, all association analyses are exploratory and hypothesis-generating. The Benjamini–Hochberg procedure will be applied within each family of tests, defined by questionnaire section, to control the false discovery rate [25]; both unadjusted and adjusted p-values will be reported.

The internal consistency of the four thematic sections, initially assessed in the pilot study (Section 2.9), will be reassessed in the main study sample using Cronbach’s alpha. In addition, the dimensional structure of the 21 five-point ordinal items (items 12–25 and 30–36) will be examined using exploratory factor analysis based on polychoric correlations, after reverse-coding items 12 and 22, which are worded in the opposite direction; with 385 respondents, this corresponds to approximately 18 respondents per item. Items 26–29, which use different response formats, and the open-ended item 37 will not be included. Sampling adequacy will be assessed using the Kaiser–Meyer–Olkin measure and Bartlett’s test, the number of factors will be determined by parallel analysis, and an oblique rotation will be used. Composite scores will be calculated only for factors supported by these analyses, with Cronbach’s alpha and McDonald’s omega reported for each; analyses using composite scores will be reported separately from item-level analyses. As these analyses will use the same sample as the substantive analyses, confirmatory factor analysis in an independent sample will be needed before the instrument can be considered fully validated.

The number of submitted questionnaires will be reported in relation to the numbers of patients screened, eligible and enrolled (Section 2.4).

To help readers interpret the findings, the main-study report will describe the meteorological and air-quality conditions in Bucharest during the enrolment period at aggregate level (for example, monthly mean temperatures, the number of heatwave and cold-spell days, and mean PM_2.5 and PM₁₀ concentrations from the national monitoring networks). These data will not be linked to individual responses, because the questionnaire asks about patients’ general experience rather than about a specific date; the study remains perception-based.

All statistical tests will be two-sided, with a p-value <0.05 considered statistically significant (for association analyses, after the false-discovery-rate adjustment). The statistical software used for the analyses will be specified in the final manuscript.

3. Results

3.1. Expected Contribution

This study is designed to characterize how patients with cardiovascular disease perceive and respond to extreme temperatures, severe weather events, and air pollution. Although these exposures are clinically relevant [5,6], existing surveys have primarily examined environmental awareness and attitudes in the general population or specific occupational and educational groups [7,10,11,12,14]. The present study addresses this gap by focusing on hospitalised cardiovascular patients and examining perceived health effects, emotional responses, protective behaviours, healthcare use, and communication with physicians within a single questionnaire. Evidence linking air pollution to peripheral artery disease [8], supports the inclusion of patients with any disease of the circulatory system, although the questionnaire cannot distinguish perceptions by vascular territory.

3.2. Comparison with Existing Surveys

Previous surveys have identified substantial awareness of environmental health risks alongside important gaps in specific knowledge. In a German survey of 1000 adults, 51% were very or extremely concerned that climate change would affect their health, 87% considered heat waves potentially harmful to people with chronic illnesses, and 40% wished to receive advice from their general practitioner about personal health risks and protection during heat waves [5]. In Ireland, 66.5% of 1005 respondents recognized air pollution as a national problem, compared with 35.1% who recognized it as a local problem; fewer than half identified particulate matter as a harmful pollutant [10]. Similarly, among 1092 Polish adults, 90.5% agreed that air pollution can cause serious health problems, although knowledge of specific pollutants and their sources remained incomplete [11]. At European level, the proportion of respondents identifying both health and climate change as top national priorities was only 1.2% in 2023 and 2024, suggesting that these issues are often considered separately [2]. This finding concerns public priorities rather than knowledge of specific health risks, but it highlights the challenge of communicating climate change in terms of its health implications. Items 13–18 of the present questionnaire specifically ask patients whether heat, cold, and air pollution worsen their own cardiac symptoms, addressing a patient-centred dimension not directly assessed in this form by the general-population surveys discussed above.

The emotional dimension of environmental risk perception has also been documented. In Thessaloniki, approximately 90% of 245 respondents reported some degree of anxiety related to air pollution [12]. Among 352 Colombian medical students, 25.9% considered air pollution a multifactorial problem on which individual actions were ineffective [7]. Although these populations differ from hospitalised cardiovascular patients, their findings support examining both emotional responses and perceived personal control. Section 3 of the present questionnaire addresses stress, helplessness, quality-of-life implications, and perceived control, allowing these dimensions to be explored in a clinically relevant population.

Knowledge does not necessarily translate into protective behaviour. Among Colombian medical students, knowledge scores were relatively high, whereas practice scores were lower, and regression analysis found no association between knowledge and practices [7]. Among 3062 members of the Indian healthcare workforce, 80.9% were aware of the health effects of heat and 79.2% of those of cold [14]. These results suggest that awareness alone may not ensure protective action, although differences in populations and questionnaire instruments limit direct comparisons. By combining self-rated information (items 30–31) with questions on protective behaviours, treatment-related responses, and healthcare use (items 23–29), the present study may help describe the relationship between perceived knowledge and reported actions among cardiovascular patients. Any associations identified will remain cross-sectional and cannot establish causality.

Physician communication represents another relevant aspect. In the German survey, 40% of respondents wished to receive advice from their general practitioner about climate-related health risks and heat protection [5], whereas in Thessaloniki approximately 20% reported having received advice from a physician about air pollution [12]. Item 33 of the present questionnaire assesses whether physicians inform patients about environmental influences on health, while item 36 addresses interest in educational programmes. These items may help identify perceived gaps in clinical communication and inform discussion of environmental health counselling in cardiovascular practice.

3.3. Heat, Cold and the Local Context

Bucharest provides a relevant setting for examining patients’ perceptions of environmental cardiovascular risks, as its continental climate is characterized by cold winters and hot summers. Both high and low temperatures have been associated with cardiovascular mortality in Bucharest and its surrounding rural areas, with differences in vulnerability across population groups [26,27]. The relevance of heat exposure is further illustrated by a 2025 attribution analysis covering 854 European cities, which identified Bucharest among the capitals with the highest estimated age-standardised excess heat-related mortality rates that summer [3]. These figures are model-based estimates rather than counts of individually verified deaths, and they do not establish which local factors account for differences between cities.

Air pollution represents an additional environmental concern. A national assessment of particulate matter concentrations in Romania during 2009–2018 documented the health risks associated with PM_2.5 and PM₁₀ exposure, including cardiopulmonary effects, and highlighted the relevance of these pollutants in the Bucharest region [28]. More recently, the 2026 Europe Cities report of the Lancet Countdown identified the residential sector as the dominant contributor to PM_2.5-attributable mortality in most European cities; in 2023, PM_2.5-attributable mortality associated with solid-fuel use was more than three times higher in eastern than in northern European cities [4]. Against this background, the present study examines how patients with cardiovascular disease perceive the effects of heat, cold, and air pollution on their cardiac symptoms and daily lives, and explores their reported protective behaviours, healthcare-related responses, and information needs. The local environmental context provides the rationale for investigating these patient-reported experiences, rather than estimating individual exposure or quantifying the clinical effects of environmental stressors.

4. Discussion

4.1. Strengths and Limitations

A principal strength is the development of a questionnaire specifically intended for patients with cardiovascular disease, integrating their perceptions of heat, cold, and air pollution with psychological impact, protective behaviours, healthcare use, and information needs. The questionnaire underwent multidisciplinary content development involving expertise in cardiology, clinical psychology, and environmental physics, followed by independent expert assessment, which provided preliminary evidence of content validity within a three-member panel (S-CVI/Ave = 1), and pilot testing in 39 participants. The pilot study also provided preliminary evidence of acceptable-to-good internal consistency for the four thematic sections. Additional strengths include consecutive recruitment of eligible patients and verification of cardiovascular diagnoses using medical records. Furthermore, offering assisted completion on a study phone may reduce barriers related to digital literacy and access, helping to include patients who might otherwise be less likely to participate in an online survey [12].

Several limitations should be acknowledged. First, the single-centre design and recruitment of hospitalised patients may limit generalisability to outpatients, community-dwelling patients, and other geographical settings. Hospitalised patients may differ from outpatients and community-dwelling patients in disease severity, comorbidities, socioeconomic circumstances and risk perception, and a recent acute event may increase the salience of perceived environmental triggers; the findings will therefore apply to hospitalised patients and cannot be generalised to all patients with cardiovascular disease. Second, consecutive recruitment is a non-probability sampling approach; therefore, despite the planned sample size, the sample may not be representative of all patients with cardiovascular disease in Bucharest or Romania. Third, the questionnaire relies on self-reported and retrospective information, which may be influenced by recall bias, social desirability bias and recent environmental events, and self-reported comorbidities (item 8) may be misclassified. Items 26–28 have no defined recall period, and time since diagnosis is not recorded, so respondents with longer disease duration have had more opportunity to experience these events. The study measures perceived exposure and perceived symptom aggravation, but not actual exposure. In the absence of objective environmental measurements and clinical outcome data, reported symptom aggravation reflects the patient’s attribution rather than a verified temporal or causal relationship, and causal relationships cannot be established. Fourth, some items (items 16, 18, 21, 23–28 and 32–35) refer jointly to severe weather and air pollution, limiting the ability to distinguish the perceived effects of individual environmental exposures. Fifth, item 26 does not distinguish physician-directed from self-initiated treatment changes, categories 2 and 3 of item 29 (considering a change and having thought about it without acting) are not clearly distinct, and, in the absence of a “don’t know” option, some neutral responses may reflect an inability to judge rather than a considered opinion. Sixth, the dimensional structure of the instrument has not yet been established, and the pilot internal-consistency coefficients are preliminary. Seventh, the broad ICD-10 definition of cardiovascular disease results in a heterogeneous population; because specific diagnoses are not recorded, differences in environmental vulnerability between conditions may dilute the associations observed in the overall sample. Eighth, several factors that may influence both environmental vulnerability and risk perception are captured only partially (socioeconomic status through education and occupation; residential environment through urban or rural residence) or not at all (income, housing conditions, and access to air conditioning or heating), so residual confounding cannot be excluded. Future versions of the questionnaire should consider separate items for each exposure, a defined recall period, a question on who initiated treatment changes, a “don’t know” option, recording of exact age, and items on income, housing and access to cooling and heating.

4.2. Implications

The report of the main study will include a recruitment flow diagram, in the style of the CONSORT diagram and as recommended by the STROBE statement for observational studies [29], showing the numbers of patients screened, eligible, invited and enrolled, the reasons for ineligibility and non-participation, and the number of submitted questionnaires (Section 2.4). The study may identify topics requiring greater attention in patient education, including recognition of environmental health risks, reported protective practices, and access to physician counselling. Its findings could inform the development of educational materials and communication strategies addressing heat, cold, severe weather, and air pollution in cardiovascular care. These applications should be guided by the observed results and interpreted in light of the study’s single-centre design and reliance on self-reported data. In a subsequent phase, a revised version of the questionnaire, incorporating the changes recommended in Section 4.1, including a defined recall period, could be implemented in a multicentre, nationwide study to assess regional variations in patients’ perceptions, protective behaviours, and information needs, and to evaluate the generalisability of the findings across Romania.

5. Conclusions

This protocol describes a single-centre, cross-sectional study using a purpose-built questionnaire with preliminary evidence of content validity, pretested in a pilot sample, to investigate perceptions and reported responses to environmental stressors among hospitalised patients with cardiovascular disease in Bucharest. By combining questions on perceived health effects, emotional responses, protective behaviours, and physician communication, the study aims to provide clinically relevant information that may support patient education and environmental health counselling in cardiovascular practice. The findings will describe patients’ perceptions and reported practices; they will not establish causal effects of environmental exposures on cardiovascular health. Further psychometric evaluation in the main study sample will be required before the instrument is used more widely.