Int J Biometeorol DOI 10.1007/s00484-014-0831-7
SHORT NOTE
The influence of weather on health-related help-seeking behavior of senior citizens in Hong Kong Ho Ting Wong & Marcus Yu Lung Chiu & Cynthia Sau Ting Wu & Tsz Cheung Lee & Senior Citizen Home Safety Association
Received: 7 January 2014 / Revised: 3 April 2014 / Accepted: 5 April 2014 # ISB 2014
Abstract It is believed that extreme hot and cold weather has a negative impact on general health conditions. Much research focuses on mortality, but there is relatively little community health research. This study is aimed at identifying high-risk groups who are sensitive to extreme weather conditions, in particular, very hot and cold days, through an analysis of the health-related help-seeking patterns of over 60,000 Personal Emergency Link (PE-link) users in Hong Kong relative to weather conditions. In the study, 1,659,716 PE-link calls to the help center were analyzed. Results showed that females, older elderly, people who did not live alone, non-subsidized (relatively high-income) users, and those without medical histories of heart disease, hypertension, stroke, and diabetes were more sensitive to extreme weather condition. The results suggest that using official government weather forecast reports to predict health-related help-seeking behavior is feasible. An evidence-based strategic plan could be formulated by using a method similar to that used in this study to identify
H. T. Wong Department of Social Work, Hong Kong Baptist University, Hong Kong, SAR, China e-mail: [email protected] M. Y. L. Chiu (*) Department of Social Work, Faculty of Arts & Social Sciences, National University of Singapore, Block AS3, Level 4, 3 Arts Link, Singapore 117570, Singapore e-mail: [email protected] C. S. T. Wu School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, SAR, China e-mail: [email protected] T. C. Lee Hong Kong Observatory, 134A, Nathan Road, Tsim Sha Tsui, Kowloon, Hong Kong, China e-mail: [email protected]
high-risk groups. Preventive measures could be established for protecting the target groups when extreme weather conditions are forecasted. Keywords Chinese . Emergency . Telehealth . Weather . Help-seeking . Weather forecast
Introduction In recent decades, against the background of observed weather and climate change, there is a growing concern about the threats of extreme weather events, including extremely hot and cold conditions, to public health (IPCC 2013). While the mortality rate is one of the prime concerns for studying the connection between weather and health, relevant studies are, to a certain extent, limited by the relatively few cases of death that can be directly attributed to weather events and the lack of a comprehensive database. Studies have used data from more common events such as hospital admission to assess health risks (Goggins et al. 2012; Chan et al. 2013). Moreover, data from telehealth systems is also a useful alternative for analyzing the response and help-seeking patterns of specific user groups and provides a more comprehensive picture of the effects of weather on our daily life than does a study on mortality rates alone. Generally speaking, most studies suggest that the mortality rate or hospital admission rate will increase when the temperature is higher (lower) than a threshold value in hot (cold) seasons. The studies of Chan et al. (2013) and Leung et al. (2008) also indicated that the elderly age group may be more vulnerable to temperature changes than other age groups. The proportion of the Hong Kong population that is senior citizens has increased continuously over the past decade. The latest projection from the Hong Kong SAR Government suggests that the number of elderly people will increase to about 2.56
Int J Biometeorol
million by 2041, about 30 % of the total population (HKSAR 2013). The growth of the elderly population will inevitably result in a surge in the need for healthcare and related services, particularly during extreme weather conditions. The Personal Emergency Link (PE-link), a well-known telehealth system in Hong Kong, is one of the core services of Senior Citizen Home Safety Association (SCHSA) for elderly citizens. Currently, SCHSA has more than 80,000 active users of the PE-link in Hong Kong. Similar telehealth systems exist in other major cities in China, including Shanghai, Guangzhou, Shenzhen, and Macau. PE-link is an indoor remote trigger connected to a 24-h call center. When users need assistance, they can communicate with call center operators by simply pressing a remote trigger. If necessary, an operator will help the user call for ambulance services for hospitalization or other emergency assistance, and the call record will be marked according to the purpose (SCHSA 2013). Given the complete data records and the popularity of the PE-link service, research studies have been conducted to analyze the effects of weather on PE-link callers in an effort to quantify the effects of weather on community health in Hong Kong. For example, Chan et al. (2011) analyzed the PE-link call records for the warm seasons from 1998 to 2006 and found that females, elderly aged ≤75 years, and those living alone were more sensitive to the effects of high temperatures. Mok and Leung (2009) found that PE-link users were affected by both hot and cold weather. PE-Link users were more vulnerable to low temperatures in cold seasons than high temperatures in hot seasons. Also, the health risk to senior citizens is higher under dry and cold weather conditions. With a view to better understanding the effect of weather on the demands of PE-link, this study further analyzed the relationship between the number of PE-Link calls which required subsequent hospitalization during both hot and cold conditions for different subgroup users. The subgroups that are more sensitive to weather conditions were identified to help formulate more focused and specific assistance for the elderly community during extreme weather conditions.
PE-link calls, which required subsequent hospitalization, were first selected from the master database and aggregated into time series data of the daily number of PE-link calls which required subsequent hospitalization. In order to eliminate the effects of the annual increase of new PE-link users, the data series was further normalized into the number of daily PE-link calls which required subsequent hospitalization per 50,000 users. The 3day moving average2 was applied to the data series to remove random noise, and 4-day time lag3 data was used to address the time lag of weather effects (Wong and Lai 2012). Afterward, multiple linear regression models were developed between the daily number of PE-link calls which required subsequent hospitalization and the meteorological time series data for different subgroups by using SPSS. The meteorological variables considered in this study included average temperature (T), difference of average temperature between two consecutive days, average relative humidity (RH), average air pressure, and interaction between T and RH. SPSS built-in variable selection function was used for model development for attempting to include only the optimum set of variables in each regression model (Chan 2003). As the adjusted R2 obtained from the regression analysis in this study represents the percentage of variance in the daily number of PE-link calls which required subsequent hospitalization that can be explained by meteorological variables (Chan 2003) and based on successful experience in similar research (Yan 2000a, b; Wong and Lai 2012), it was used to compare the sensitivity of the weather effects among different subgroups of PE-link users.
Data and methods
Results
Data
Table 1 shows that female users had a higher adjusted R2 than male users. In general, older users also had a higher adjusted
One million six hundred fifty-nine thousand seven hundred sixteen PE-link call records during the 3-year period from May 2006 to April 2009 were analyzed in this study. Each record contained information about the user: gender, age, living alone (yes/no), subsidized (yes/no), heart disease history (yes/no), hypertension history (yes/no), stroke history (yes/ no), diabetes history (yes/no), and whether subsequent hospitalization was required (yes/no).1
For the meteorological data, observations made at the Hong Kong Observatory (HKO) located in the urban center of Hong Kong were used. They included daily average temperatures, relative humidity, and air pressure. The data about sunshine, rainfall, and wind speed were not used in the study because most PE-Link users remain indoors when contacting the call center (Mok and Leung 2009). Data analysis
1 The proportion of PE-link calls which required subsequent hospitalization was 5.6 %. 2 The parameters were set by testing the association between PE-Link demand and the corresponding air temperature data series (Leung et al. 2008; Wong and Lai 2012). 3 The parameters were set by testing the association between PE-Link demand and the corresponding air temperature data series (Leung et al. 2008; Wong and Lai 2012).
Int J Biometeorol Table 1 The adjusted R2 obtained from the regression model for different subgroups Target group
Number Adjusted R2 Adjusted R2 difference of usersa
Gender
20,855 41,102 3,055 14,221 32,033 12,654 29,602 32,361
0.278 0.332 0.013 0.110 0.330 0.270 0.264 0.352
34,696 27,214 13,516 48,447 29,290 32,673 5,917 56,046 14,577 47,386 61,963
0.295 0.354 0.205 0.381 0.251 0.371 0.063 0.420 0.249 0.360 0.415
Male Female Age 84 Living status Living alone Not living alone Subsidized users Yes No Heart disease Yes history No Hypertension Yes history No Stroke history Yes No Diabetes history Yes No Overall a
Figures as of 30 April 2009
b
Reference group
0.054
0.097 0.317 0.257 0.088
0.059 0.176 0.12 0.357 0.111
R2, although the oldest group had a slightly lower adjusted R2 compared to users aged 75–84. Users not living alone and non-subsidized users also had a higher adjusted R2. Regarding medical histories, users without a history of heart disease, hypertension, stroke, and diabetes had a higher adjusted R2. Number of the PE-link users under different subgroups is also listed in Table 1.
Discussion There is a general perception that female, older elderly, those living alone, subsidized users (i.e., less income), and users with the selected medical histories are more prone to the impact of weather. Based on subgroup analysis of the number of PE-Link calls that required subsequent hospitalization, the present study suggests that the study results of gender and age were consistent with the said common perception. This is generally in line with corresponding findings reported by Chan et al. (2011). However, our study revealed that PE-Link users in the subgroups of non-subsidized, not living alone, and without medical histories (especially heart disease and stroke) may be more sensitive to weather effects than users who were subsidized, living alone, and had medical histories. It is noteworthy that, consistent with other healthcare studies, utilization of services is characterized not by those who had less resources
and support, but by those who are more health conscious and who dare to bring down the psychological barrier to make use of available help. To further investigate the reasons behind these response patterns, PE-link service frontline staff members in the call center were interviewed, and several possible factors that were not depicted in the analyzed dataset were identified: 1. A large number of users who do not live alone were frail elderly who required family members or domestic workers to live with them and take care of them. Their frailty may explain why this subgroup is more sensitive to weather extremes. 2. Users with medical histories of heart disease or stroke were found to be relatively less sensitive to weather when compared to those without medical histories. This may be because they were more concerned about their health status given their medical history and proactively took preventative measures to protect themselves against the possible impacts of extreme weather temperatures. Moreover, the current strategy of PE-link service is to regularly remind those users with select medical history to take care of themselves before and during very hot and cold days. This tactic might increase conscious awareness among users about the health impacts of extreme weather conditions and thereby reduce the chance of hospital admission. Since it would be very resource and man power demanding in visiting all PE-link users within a short period of time, the above findings are particularly useful for SCHSA to design their intervention programs to enhance the PE-link services and to promote public awareness on the caring of elderly. For example, based on the weather forecast 1 week ahead, SCHSA can arrange volunteers to carry out timely home visits or telephone calls to remind the high-risk groups about the impending cold surges or heat waves and to identify their needs. As a step forward, based on the validated regression model for the relationship between weather and PE-link demand, PE-link service providers could also consider developing a demand forecast model using HKO’s 7-day weather forecast to prioritize their work and better utilize the available resources (Wong and Lai 2014). This is particularly useful for launching preventive measures a few days before the occurrence of the extremely hot or cold weather. Moreover, with a rapidly aging population and against the background of climate change, the effects of extreme weather on human health could become more significant in the coming decades. The health-related help-seeking patterns generated from the relationship between weather and PE-link demand can be adopted to project the plausible changes in the demand in the next few decades by using the population projection data (Lai and Wong 2014). Such projections will be useful for timely planning of the resource needed in a foreseeable future.
Int J Biometeorol
Conclusion To conclude, this study filled the research gap by identifying high-risk groups who may be affected by extreme weather conditions, in particular, cold and very hot days, from the community health perspective. PE-link service providers can use this information as a basis for launching tailor-made intervention programs targeted at the high-risk groups and planning their elderly caring work before the arrival of the cold surge or heat wave. However, the identified high-risk groups were out of common expectations, and the identified high-risk groups may only represent users who were more health conscious but not the most vulnerable group. Hence, it is recommended that further research by using different research methods (e.g., qualitative and mixed methods) could be conducted to study if the high-risk groups found represent the most vulnerable group in the population. Moreover, a larger-scale study can be conducted with more data and that possible confounding factors could be controlled so that the results of this study can be further validated.
References Chan YH (2003) Biostatistics 201: linear regression analysis. Singap Med J 44:280–285 Chan EYY, Goggins WB, Kim JJ, Griffiths SM, Ma T (2011) Helpseeking behavior during elevated temperature in Chinese population. J Urban Health 88:637–650
Chan EYY, Goggins WB, Yue JSK, Lee P (2013) Hospital admissions as a function of temperature, other weather phenomena and pollution levels in an urban setting in China. Bull World Health Organ 91: 576–584 Goggins WB, Woo J, Ho S, Chan EYY, Chau PH (2012) Weather, season, and daily stroke admissions in Hong Kong. Int J Biometeorol 56: 865–872 HKSAR (2013) Embracing the Challenges Ahead, 2013–14 Budget Speech. Available at: http://www.budget.gov.hk/2013/eng/ budget27.html. Accessed 7 Jan 2014 IPCC (2013) Summary for policymakers. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge Lai PC, Wong HT (2014) Weather and age–gender effects on the projection of future emergency ambulance demand in Hong Kong. Asia Pac J Public Health. doi:10.1177/1010539512460570 Leung YK, Yip KM, Yeung KH (2008) Relational between thermal index and mortality in Hong Kong. Meteorol Appl 15:339–408 Mok HY, Leung B (2009) The impact of cold and hot weather on senior citizens in Hong Kong. HKMetS Bull 19:9–31 SCHSA (2013) Senior Citizens Home Safety Association Annual Report 2011–2012. http://schsa.org.hk/filemanager/tc/content_70/SCHS A_Annual_report_11-12_eachpage.pdf. Accessed 7 Jan 2014 Wong HT, Lai PC (2012) Weather inference and daily demand for emergency ambulance services in Hong Kong. Emerg Med J 29: 60–64 Wong HT, Lai PC (2014) Weather factors in the short-term forecasting of daily ambulance calls. Int J Biometeorol. doi:10.1007/s00484-0130647-x Yan YY (2000a) The influence of weather on human mortality in Hong Kong. Soc Sci Med 50:419–427 Yan YY (2000b) The influence of weather on suicide in Hong Kong. Percept Mot Skills 91:571–577
SHORT NOTE
The influence of weather on health-related help-seeking behavior of senior citizens in Hong Kong Ho Ting Wong & Marcus Yu Lung Chiu & Cynthia Sau Ting Wu & Tsz Cheung Lee & Senior Citizen Home Safety Association
Received: 7 January 2014 / Revised: 3 April 2014 / Accepted: 5 April 2014 # ISB 2014
Abstract It is believed that extreme hot and cold weather has a negative impact on general health conditions. Much research focuses on mortality, but there is relatively little community health research. This study is aimed at identifying high-risk groups who are sensitive to extreme weather conditions, in particular, very hot and cold days, through an analysis of the health-related help-seeking patterns of over 60,000 Personal Emergency Link (PE-link) users in Hong Kong relative to weather conditions. In the study, 1,659,716 PE-link calls to the help center were analyzed. Results showed that females, older elderly, people who did not live alone, non-subsidized (relatively high-income) users, and those without medical histories of heart disease, hypertension, stroke, and diabetes were more sensitive to extreme weather condition. The results suggest that using official government weather forecast reports to predict health-related help-seeking behavior is feasible. An evidence-based strategic plan could be formulated by using a method similar to that used in this study to identify
H. T. Wong Department of Social Work, Hong Kong Baptist University, Hong Kong, SAR, China e-mail: [email protected] M. Y. L. Chiu (*) Department of Social Work, Faculty of Arts & Social Sciences, National University of Singapore, Block AS3, Level 4, 3 Arts Link, Singapore 117570, Singapore e-mail: [email protected] C. S. T. Wu School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, SAR, China e-mail: [email protected] T. C. Lee Hong Kong Observatory, 134A, Nathan Road, Tsim Sha Tsui, Kowloon, Hong Kong, China e-mail: [email protected]
high-risk groups. Preventive measures could be established for protecting the target groups when extreme weather conditions are forecasted. Keywords Chinese . Emergency . Telehealth . Weather . Help-seeking . Weather forecast
Introduction In recent decades, against the background of observed weather and climate change, there is a growing concern about the threats of extreme weather events, including extremely hot and cold conditions, to public health (IPCC 2013). While the mortality rate is one of the prime concerns for studying the connection between weather and health, relevant studies are, to a certain extent, limited by the relatively few cases of death that can be directly attributed to weather events and the lack of a comprehensive database. Studies have used data from more common events such as hospital admission to assess health risks (Goggins et al. 2012; Chan et al. 2013). Moreover, data from telehealth systems is also a useful alternative for analyzing the response and help-seeking patterns of specific user groups and provides a more comprehensive picture of the effects of weather on our daily life than does a study on mortality rates alone. Generally speaking, most studies suggest that the mortality rate or hospital admission rate will increase when the temperature is higher (lower) than a threshold value in hot (cold) seasons. The studies of Chan et al. (2013) and Leung et al. (2008) also indicated that the elderly age group may be more vulnerable to temperature changes than other age groups. The proportion of the Hong Kong population that is senior citizens has increased continuously over the past decade. The latest projection from the Hong Kong SAR Government suggests that the number of elderly people will increase to about 2.56
Int J Biometeorol
million by 2041, about 30 % of the total population (HKSAR 2013). The growth of the elderly population will inevitably result in a surge in the need for healthcare and related services, particularly during extreme weather conditions. The Personal Emergency Link (PE-link), a well-known telehealth system in Hong Kong, is one of the core services of Senior Citizen Home Safety Association (SCHSA) for elderly citizens. Currently, SCHSA has more than 80,000 active users of the PE-link in Hong Kong. Similar telehealth systems exist in other major cities in China, including Shanghai, Guangzhou, Shenzhen, and Macau. PE-link is an indoor remote trigger connected to a 24-h call center. When users need assistance, they can communicate with call center operators by simply pressing a remote trigger. If necessary, an operator will help the user call for ambulance services for hospitalization or other emergency assistance, and the call record will be marked according to the purpose (SCHSA 2013). Given the complete data records and the popularity of the PE-link service, research studies have been conducted to analyze the effects of weather on PE-link callers in an effort to quantify the effects of weather on community health in Hong Kong. For example, Chan et al. (2011) analyzed the PE-link call records for the warm seasons from 1998 to 2006 and found that females, elderly aged ≤75 years, and those living alone were more sensitive to the effects of high temperatures. Mok and Leung (2009) found that PE-link users were affected by both hot and cold weather. PE-Link users were more vulnerable to low temperatures in cold seasons than high temperatures in hot seasons. Also, the health risk to senior citizens is higher under dry and cold weather conditions. With a view to better understanding the effect of weather on the demands of PE-link, this study further analyzed the relationship between the number of PE-Link calls which required subsequent hospitalization during both hot and cold conditions for different subgroup users. The subgroups that are more sensitive to weather conditions were identified to help formulate more focused and specific assistance for the elderly community during extreme weather conditions.
PE-link calls, which required subsequent hospitalization, were first selected from the master database and aggregated into time series data of the daily number of PE-link calls which required subsequent hospitalization. In order to eliminate the effects of the annual increase of new PE-link users, the data series was further normalized into the number of daily PE-link calls which required subsequent hospitalization per 50,000 users. The 3day moving average2 was applied to the data series to remove random noise, and 4-day time lag3 data was used to address the time lag of weather effects (Wong and Lai 2012). Afterward, multiple linear regression models were developed between the daily number of PE-link calls which required subsequent hospitalization and the meteorological time series data for different subgroups by using SPSS. The meteorological variables considered in this study included average temperature (T), difference of average temperature between two consecutive days, average relative humidity (RH), average air pressure, and interaction between T and RH. SPSS built-in variable selection function was used for model development for attempting to include only the optimum set of variables in each regression model (Chan 2003). As the adjusted R2 obtained from the regression analysis in this study represents the percentage of variance in the daily number of PE-link calls which required subsequent hospitalization that can be explained by meteorological variables (Chan 2003) and based on successful experience in similar research (Yan 2000a, b; Wong and Lai 2012), it was used to compare the sensitivity of the weather effects among different subgroups of PE-link users.
Data and methods
Results
Data
Table 1 shows that female users had a higher adjusted R2 than male users. In general, older users also had a higher adjusted
One million six hundred fifty-nine thousand seven hundred sixteen PE-link call records during the 3-year period from May 2006 to April 2009 were analyzed in this study. Each record contained information about the user: gender, age, living alone (yes/no), subsidized (yes/no), heart disease history (yes/no), hypertension history (yes/no), stroke history (yes/ no), diabetes history (yes/no), and whether subsequent hospitalization was required (yes/no).1
For the meteorological data, observations made at the Hong Kong Observatory (HKO) located in the urban center of Hong Kong were used. They included daily average temperatures, relative humidity, and air pressure. The data about sunshine, rainfall, and wind speed were not used in the study because most PE-Link users remain indoors when contacting the call center (Mok and Leung 2009). Data analysis
1 The proportion of PE-link calls which required subsequent hospitalization was 5.6 %. 2 The parameters were set by testing the association between PE-Link demand and the corresponding air temperature data series (Leung et al. 2008; Wong and Lai 2012). 3 The parameters were set by testing the association between PE-Link demand and the corresponding air temperature data series (Leung et al. 2008; Wong and Lai 2012).
Int J Biometeorol Table 1 The adjusted R2 obtained from the regression model for different subgroups Target group
Number Adjusted R2 Adjusted R2 difference of usersa
Gender
20,855 41,102 3,055 14,221 32,033 12,654 29,602 32,361
0.278 0.332 0.013 0.110 0.330 0.270 0.264 0.352
34,696 27,214 13,516 48,447 29,290 32,673 5,917 56,046 14,577 47,386 61,963
0.295 0.354 0.205 0.381 0.251 0.371 0.063 0.420 0.249 0.360 0.415
Male Female Age 84 Living status Living alone Not living alone Subsidized users Yes No Heart disease Yes history No Hypertension Yes history No Stroke history Yes No Diabetes history Yes No Overall a
Figures as of 30 April 2009
b
Reference group
0.054
0.097 0.317 0.257 0.088
0.059 0.176 0.12 0.357 0.111
R2, although the oldest group had a slightly lower adjusted R2 compared to users aged 75–84. Users not living alone and non-subsidized users also had a higher adjusted R2. Regarding medical histories, users without a history of heart disease, hypertension, stroke, and diabetes had a higher adjusted R2. Number of the PE-link users under different subgroups is also listed in Table 1.
Discussion There is a general perception that female, older elderly, those living alone, subsidized users (i.e., less income), and users with the selected medical histories are more prone to the impact of weather. Based on subgroup analysis of the number of PE-Link calls that required subsequent hospitalization, the present study suggests that the study results of gender and age were consistent with the said common perception. This is generally in line with corresponding findings reported by Chan et al. (2011). However, our study revealed that PE-Link users in the subgroups of non-subsidized, not living alone, and without medical histories (especially heart disease and stroke) may be more sensitive to weather effects than users who were subsidized, living alone, and had medical histories. It is noteworthy that, consistent with other healthcare studies, utilization of services is characterized not by those who had less resources
and support, but by those who are more health conscious and who dare to bring down the psychological barrier to make use of available help. To further investigate the reasons behind these response patterns, PE-link service frontline staff members in the call center were interviewed, and several possible factors that were not depicted in the analyzed dataset were identified: 1. A large number of users who do not live alone were frail elderly who required family members or domestic workers to live with them and take care of them. Their frailty may explain why this subgroup is more sensitive to weather extremes. 2. Users with medical histories of heart disease or stroke were found to be relatively less sensitive to weather when compared to those without medical histories. This may be because they were more concerned about their health status given their medical history and proactively took preventative measures to protect themselves against the possible impacts of extreme weather temperatures. Moreover, the current strategy of PE-link service is to regularly remind those users with select medical history to take care of themselves before and during very hot and cold days. This tactic might increase conscious awareness among users about the health impacts of extreme weather conditions and thereby reduce the chance of hospital admission. Since it would be very resource and man power demanding in visiting all PE-link users within a short period of time, the above findings are particularly useful for SCHSA to design their intervention programs to enhance the PE-link services and to promote public awareness on the caring of elderly. For example, based on the weather forecast 1 week ahead, SCHSA can arrange volunteers to carry out timely home visits or telephone calls to remind the high-risk groups about the impending cold surges or heat waves and to identify their needs. As a step forward, based on the validated regression model for the relationship between weather and PE-link demand, PE-link service providers could also consider developing a demand forecast model using HKO’s 7-day weather forecast to prioritize their work and better utilize the available resources (Wong and Lai 2014). This is particularly useful for launching preventive measures a few days before the occurrence of the extremely hot or cold weather. Moreover, with a rapidly aging population and against the background of climate change, the effects of extreme weather on human health could become more significant in the coming decades. The health-related help-seeking patterns generated from the relationship between weather and PE-link demand can be adopted to project the plausible changes in the demand in the next few decades by using the population projection data (Lai and Wong 2014). Such projections will be useful for timely planning of the resource needed in a foreseeable future.
Int J Biometeorol
Conclusion To conclude, this study filled the research gap by identifying high-risk groups who may be affected by extreme weather conditions, in particular, cold and very hot days, from the community health perspective. PE-link service providers can use this information as a basis for launching tailor-made intervention programs targeted at the high-risk groups and planning their elderly caring work before the arrival of the cold surge or heat wave. However, the identified high-risk groups were out of common expectations, and the identified high-risk groups may only represent users who were more health conscious but not the most vulnerable group. Hence, it is recommended that further research by using different research methods (e.g., qualitative and mixed methods) could be conducted to study if the high-risk groups found represent the most vulnerable group in the population. Moreover, a larger-scale study can be conducted with more data and that possible confounding factors could be controlled so that the results of this study can be further validated.
References Chan YH (2003) Biostatistics 201: linear regression analysis. Singap Med J 44:280–285 Chan EYY, Goggins WB, Kim JJ, Griffiths SM, Ma T (2011) Helpseeking behavior during elevated temperature in Chinese population. J Urban Health 88:637–650
Chan EYY, Goggins WB, Yue JSK, Lee P (2013) Hospital admissions as a function of temperature, other weather phenomena and pollution levels in an urban setting in China. Bull World Health Organ 91: 576–584 Goggins WB, Woo J, Ho S, Chan EYY, Chau PH (2012) Weather, season, and daily stroke admissions in Hong Kong. Int J Biometeorol 56: 865–872 HKSAR (2013) Embracing the Challenges Ahead, 2013–14 Budget Speech. Available at: http://www.budget.gov.hk/2013/eng/ budget27.html. Accessed 7 Jan 2014 IPCC (2013) Summary for policymakers. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge Lai PC, Wong HT (2014) Weather and age–gender effects on the projection of future emergency ambulance demand in Hong Kong. Asia Pac J Public Health. doi:10.1177/1010539512460570 Leung YK, Yip KM, Yeung KH (2008) Relational between thermal index and mortality in Hong Kong. Meteorol Appl 15:339–408 Mok HY, Leung B (2009) The impact of cold and hot weather on senior citizens in Hong Kong. HKMetS Bull 19:9–31 SCHSA (2013) Senior Citizens Home Safety Association Annual Report 2011–2012. http://schsa.org.hk/filemanager/tc/content_70/SCHS A_Annual_report_11-12_eachpage.pdf. Accessed 7 Jan 2014 Wong HT, Lai PC (2012) Weather inference and daily demand for emergency ambulance services in Hong Kong. Emerg Med J 29: 60–64 Wong HT, Lai PC (2014) Weather factors in the short-term forecasting of daily ambulance calls. Int J Biometeorol. doi:10.1007/s00484-0130647-x Yan YY (2000a) The influence of weather on human mortality in Hong Kong. Soc Sci Med 50:419–427 Yan YY (2000b) The influence of weather on suicide in Hong Kong. Percept Mot Skills 91:571–577
