Executive Summary

Executive Summary

Background and objectives: Locally derived data are needed for the SAR Government and regulatory agencies to support the drafting and implementation of effective legislation for the management and control of the environment. In 1996, the Environmental Protection Department commissioned the Chinese University of Hong Kong to carry out a study on health effects of air pollution and in 1997 the University of Hong Kong to carry out a follow-up validation study. The study performed by the University of Hong Kong identified some irregularities in the way the data were selected, managed and analysed in the previous study. It also obtained effect estimates which were more comparable to the results obtained by the collaborating centres of Air Pollution on Health: a European Approach (APHEA). Finally it pointed to new directions in methodology and in modelling methods adopted by the APHEA Phase Two study which could be used in Hong Kong.

In this study, reported here, the problems of the earlier Hong Kong studies were resolved in that (i) three years of data were used to obtain better models; (ii) non-linearity effects in the pollutant and covariates were modelled using a generalized additive modelling method; and (iii) patterns in the daily variations were explained as far as possible by using more appropriate wave lengths, introducing influenza epidemics in the models and using more recently developed statistical methods to control for autocorrelation in the data.

Health outcomes: Daily counts of hospital admissions with a discharge diagnosis of respiratory disease (ICD9 460-519), chronic obstructive pulmonary disease (COPD) (490-496 excluding 493), asthma (493), cardiac disease (390-429) and ischaemic heart disease (IHD) (410-414), for all ages, as well as hospital deaths due to respiratory (460-519) and cardiovascular (390-459) diseases, for all ages, were used as the 7 major health outcomes. Among them, admissions were sub-categorised as follows: for respiratory disease, into 0-14, 15-64, 65-74 and 75+ age groups; for asthma, into 0-14 and 15-64 age groups; for cardiac disease, into 15-64, 65-74 and 75 age groups and for IHD, into 15-64, 65-74 and 75+ age groups. These categories for all ages and specific age groups formed 19 categories of health outcomes for analysis.

Summary of findings: The effects for each pollutant without adjustment for co-pollutant, on the above-mentioned 19 health outcomes and with adjustment for the most confounding co-pollutant, on the above 7 major health outcomes are summarized in this section.

NO2: NO2 was shown to have strong effects on all the 19 health outcomes under study without adjustment for co-pollutants, except on the IHD 15-64 and all ages groups. The effects on the 7 major health outcomes, with adjustment for co-pollutant and autocorrelation, were significant at the 5% level except IHD admissions and cardiovascular hospital deaths for all ages. The effects on asthma admissions and respiratory deaths for all ages were likely to be non-linear. The effects on respiratory, COPD, cardiac and IHD admissions for all ages were higher in the cool than in the warm season.

SO2: SO2 did not have an effect on most of the 19 health outcomes under study without adjustment for co-pollutant. With adjustment for co-pollutant and autocorrelation, it had an effect on COPD admissions for all ages (p=0.005). The effect on cardiac admissions for all ages was higher in the cool than in the warm season.

RSP: RSP was associated with most of the 19 health outcomes under study without adjustment for co-pollutant. With adjustment for co-pollutants and autocorrelation, it had an effect on respiratory (p=0.023) and asthma (p=0.045) admissions for all ages and the effects were likely to be non-linear. The effect on cardiac admissions for all ages was higher in the cool than in the warm season.

O3: O3 had effects on most of the 19 health outcomes under study without adjustment for co-pollutant. With adjustment for co-pollutant and autocorrelation, it had strong effects on respiratory and asthma admissions for all ages (p=0.000). The effects on COPD, cardiac and IHD admissions for all ages were stronger in the cool than in warm season.

FSP: FSP (PM2.5) had effects on most of the 19 health outcomes under study without adjustment for co-pollutant. With adjustment for co-pollutant, the effects were not significant for all of the 7 major health outcomes for all ages (p>0.192). But as the pollutant was measured in only one monitoring station and there were missing data problems, the effects need to be further investigated.

Public health implications: The results from this study showed strong short-term effects of air pollutants on hospital admissions and hospital deaths.

In the daily number of hospital admissions for all ages, there was a 4-8% increase due to respiratory disease; 5-15% increase due to chronic obstructive pulmonary disease and 11-22% increase due to asthma. There was also a 3-7% increase in admissions due to a cardiac disease. There was a 3-5% increase due to a cardiovascular disease (ICD9 390-459), which includes cardiac disease as well as cerebrovascular disease and diseases of arteries, arterioles, capillaries, veins and lymphatics. These percentage increases are all expressed as rates per 50 ug/m3 increase in the concentration of air pollutants (NO2, SO2, O3 and RSP).

In the daily number of hospital deaths for all ages, there was a 10% increase due to respiratory disease and 8% increase due to cardiovascular disease for every 50 ug/m3 increase in NO2. For other pollutants the increases were not significant and they need to be further investigated.

The results suggest that air pollution may show stronger health effects in Hong Kong than those obtained in European cities.

Conclusions: The results from this study show that short-term effects of air pollutants on health are strong. It is both feasible and worthwhile to develop further studies on the health effects of air pollution with a longer series of data, using the methodology adopted by the APHEA protocol. Territory-wide SAR mortality, based on daily mortality counts from the Census and Statistics Department, should be included as one of the main health outcomes to be studied.

1.0 Background and introduction

In 1997, the Department was offered a contract by the Environmental Protection Department (EPD) to study the short-term effects of ambient air pollution on public health using hospital admission data in Hong Kong for the period January 1995 to June 19971 and to validate the findings of another similar study on data for the period January 1994 to June 1996.2 The results revealed that, similar to other studies overseas, there were significant effects on morbidity and mortality of individual pollutants i.e. nitrogen dioxide (NO2), sulphur dioxide (SO2), respirable suspended particulates (RSP) and ozone (O3). The study performed by the Department modified the way the data were selected, managed and analysed in the previous study and obtained effect estimates which were more comparable to the results obtained by collaborating centres of Air Pollution on Health: a European approach (APHEA). In line with new directions in methodology and in modelling methods adopted by the APHEA-2 protocol,3 our report also pointed out some limitations of both of the Hong Kong studies and indicated what should be done in any future studies.

The limitations in the Hong Kong studies will be further resolved and discussed in this new phase of study. First, with three years of data, a better prediction model will be developed. Second, non-linearity effects in the pollutant(s) or covariates will be modelled using a generalized additive model (in statistical package S-Plus). Third, unexplained patterns in the daily variations will be explained as far as possible by (i) using more appropriate wave lengths for seasonality and terms for time trends, to eliminate as far as possible all unobserved confounding effects; (ii) introducing influenza epidemics in the model to control as much of the observed confounding effects as possible in addition to those due to air temperature and humidity etc; and (iii) using more recently developed statistical methods to control for any auto-correlation and/or extra-Poisson variations in the data.

2.0 Scope and objectives

The following tasks will be completed by the Department of Community Medicine.

  1. Single pollutant: A model for each health outcome is fitted with terms to account for all long-term and seasonal patterns and other possible confounding effects. Short-term daily variations are then studied and accounted for by daily variations in SO2, NO2, RSP and O3 individually, based on data from 1 January 1995 to 31 December 1997.

  2. Multiple pollutants: From each single pollutant model, the joint effects of each pollutant with other pollutants are studied by putting them in the model. Interactions and multi-collinearity among pollutants are also studied and taken account of, based on the same data as in a) above.

  3. Single pollutant PM2.5: The effect due to PM2.5 is studied in a similar way as in a) above, but using data from 1 August 1995 to 31 December 1997 when the PM2.5 concentrations are available from the EPD.

  4. Multiple pollutants with PM2.5: Study b) above but with PM2.5 instead of RSP. Compare the models with either PM2.5 or RSP in terms of stability (reliability) in effect estimates.

  5. Verification of model: The best model developed from a) to d) above is then used to obtain the predicted daily numbers of admissions for each health outcome. The observed number of admissions in the data 1 January 1998 to 30 June 1998 are compared with the predicted for verification (i.e. predictive validity) of model.

3.0 Materials and methods

3.1

Study design

The same design as in the previous report was used. It utilized routinely collected hospital admission data, air pollutant concentration data and meteorological data by the Hospital Authority, Environmental Protection Department and Observatory. Variations in the daily number of hospital admissions were studied, and their relationships with each of the pollutants were modelled to assess the effects of air pollution on health after adjustment for confounding factors.
   
3.2

Databases

In the APHEA-2 protocol there were changes in age group definitions and missing data handling methods from those in APHEA. All the changes were incorporated in this report as far as possible. In order to be more comparable with results from APHEA studies, there were also some alterations in the choice of health outcomes, in that the category of chronic obstructive pulmonary disease (ICD9 490-496) excluded asthma (ICD9 493), and cardiac diseases (ICD9 390-429) was used instead of circulatory disease (ICD9 390-459) in the admission categories.

Hospital admission data: The hospitals included for compilation of hospital admissions were the publicly funded hospitals (accounting for 90% of hospital beds in Hong Kong) with an accident & emergency department, or those which were the referral base from the accident & emergency department of another hospital or had a 24-hour outpatient department. One other hospital which was the only hospital in the most polluted district in Hong Kong was also included. All hospitals should have a computerized system for inputting and retrieval of patient data (Basic Table A1).

Information for patients admitted to the above-mentioned hospitals between 1.1.1995 and 30.6.1998, on date of admission, discharge diagnosis (in ICD9 codes) and age (and other information not required in this study) were retrieved from the databases. The number of cases by discharge diagnosis and age, grouped according to APHEA-2 protocol (hospital deaths not in APHEA-2 categories, were also included) were summarized in Table 1 below.

Table 1: Number of patients admitted between 1.1.1995 and 30.6.1998 by discharge diagnoses and age groups

Number of cases
Discharge diagnosis (ICD9 codes) Age groups 1995-1997 1998 (Jan-June)
Respiratory disease (460-519) All ages* 223,225 46,877
0-14 yrs 74,164 16,450
15-64 yrs 48,909 9,451
65-74 yrs 38,783 7,640
75+ 61,318 13,329
Chronic obstructive pulmonary disease
(490-496 excluding 493)		 All ages* 53,374 9,795
Asthma (493) All ages* 27,080 4,286
0-14 yrs 14,005 2,133
15-64 yrs 8,520 1,332
Cardiac disease (390-429) All ages* 108,150 18,835
15-64 yrs 37,195 5,961
65-74 yrs 32,205 5,715
75+ 37,356 6,976
Ischaemic heart disease (410-414) All ages* 39,483 6,940
15-64 yrs 14,259 2,402
65-74 yrs 13,969 2,507
75+ 11,201 2,029
Deaths: Respiratory (460-519) All ages* 11,798 2,095
Cardiovascular (390-459) All ages* 11,202 1,779

* Including patients without information on age.

Pollution data: Hourly concentrations of air pollutants (Table 2) by monitoring stations (Table 3) were provided by the Air Services Group of the Environmental Protection Department.

Table 2: List of pollutants used in the study

Pollutant Unit
1. Nitrogen Dioxide {24-hr} (NO2) micrograms/cubic metre
2. Sulphur Dioxide {24-hr} (SO2) micrograms/cubic metre
3. Respirable Suspended Particulates {24-hr} (RSP)* micrograms/cubic metre
4. Ozone {8-hr} (O3) micrograms/cubic metre
5. Fine Suspended Particulates {24-hr} (FSP) micrograms/cubic metre

* Measured by Tapered Element Oscillating Microbalance (TEOM), an instrument for the continuous measurement of particulate matter in air.

Table 3: List of air pollution monitoring stations

Station Sampling height Above ground Date start operation
1. Central/West (C/W) 78m 18m (4th floor) 11/83
2. Kwai Chung (KC) 82m 25m (6th floor) 7/88
3. Kwun Tong (KT) 34m 25m (6th floor) 7/83
4. Sham Shui Po (SSP) 21m 17m (4th floor) 7/84
5. Shatin (ST) 27m 21m (5th floor) 7/91
6. Tai Po (TP) 31m 25m (6th floor) 2/90
7. Tsuen Wan (TW) 21m 17m (4th floor) 8/88

Two stations were excluded, one in Yuen Long (in suburban area) due to the extent of missing data and one in Mong Kok (in urban area) which provided concentrations measured on ground level.

In each station, for O3 the 8-hourly (9 am - 5 pm when O3 concentrations were highest) means were taken as daily data if there were at least 6 valid hourly data each day; and for the other pollutants the 24-hourly means were used as daily data if there were at least 18 valid hourly data. Other data were missing data. The percentages of non-missing daily data were shown in Table 4 below.

Table 4: Percentage of valid measures of air pollutants in each monitoring station

  Station#
Pollutant C/W KC KT SSP ST TP TW
NO2 (24-hr average)
1995 93 92 96 89 97 97 95
1996 93 96 95 90 96 95 93
1997 91 90 87 98 93 89 85
1995-97 92 93 93 92 95 93 91
1998* 97 96 94 85 97 98 96
SO2 (24-hr average)
1995 98 97 97 93 99 - 95
1996 98 95 98 89 99 - 98
1997 92 93 93 96 94 - 82
1995-97 96 95 97 93 97 - 91
1998* 97 98 94 86 97 - 94
RSP (24-h average)
1995 85 95 79 - 87 - 94
1996 99 98 92 - 87 - 86
1997 90 92 85 - 91 - 90
1995-97 91 95 85 - 89 - 90
1998* 99 95 95 - 99 - 92
O3 (8-hr average)
1995 94 93 - - - - -
1996 92 95 - - - - -
1997 87 90 - - - - -
1995-97 91 93 - - - - -
1998* 92 96 - - - - -
FSP (24-hr average)
1995 (Aug-Dec) - - - - - - 94
1996 - - - - - - 82
1997 - - - - - - 92
1995 (Aug)-97 - - - - - - 88
1998* - - - - - - 97

* the data are available only for first half year of 1998
- not available
# Abbreviations referred to Table 3.

Daily mean concentrations representing the pollutant levels in all Hong Kong areas were obtained from non-missing data for each of the pollutants under study by means of the following steps:

  1. Estimate the annual mean concentration of each pollutant in each monitoring station and subtract this annual mean from the daily concentrations of the corresponding year. The resulting series is regard as "centered".

  2. Take the arithmetic mean over all monitoring stations of these centered series day by day.

  3. Finally add the annual mean of all stations to the series obtained in step 2. The series for each health outcome is used for the analysis.

Overall, daily means obtained using APHEA-2 method were highly correlated (Spearman's correlation coefficient above 0.99) with those obtained using the previous APHEA method (Basic Table D3).

Meteorological data: Daily means of temperature and relative humidity were obtained from the Environmental Protection Department which derived the data from the Observatory.
   
3.3
 

Statistical modelling

The statistical modelling methods followed the guideline recommended by the APHEA-2 protocol and are outlined as follows:

a) Poisson regression with adjustment for over-dispersion using quasi-likelihood4 approach was used to model the hospital admission counts of the health outcomes under study.
   
b)

Using Poisson regression mentioned in a) above, a core model was established to remove all the long term trends and seasonal variations by means of the following covariates:

Covariate Explanation
t (day) Daily trend
t2 Daily curvature
cos (2?kt/365.25) Seasonality
sin (2?kt/365.25) for k cycles per year to be determined by spectral analysis
Monday Monday vs Sunday
Tuesday Tuesday vs Sunday
Wednesday Wednesday vs Sunday
Thursday Thursday vs Sunday
Friday Friday vs Sunday
Saturday Saturday vs Sunday
Holiday1 Holidays
Holiday2 Day after holidays
Temperature Linear effect of temperature
Humidity Linear effect of relative humidity
Influenza
  • Dummy variable for weeks with number of influenza admissions (ICD9 487) above 75 percentile in a year

The numbers of seasonal cycles per year were obtained from spectral analysis5 instead of 1-4 cycles used in the previous study.
   
c) The residuals from fitting the core model for each health outcome, were examined for any discernible patterns, lack of normality and autocorrelation in them.
   
d) If there are no discernible patterns and lack of normality in the residuals, then potential confounding is not evident in the daily variations. Air pollutant concentrations were then fitted to obtain the air pollution effects for pollutants. Autocorrelations were taken account of for identified models.
   
e) In order to obtain an indicator of the amount (proportion) of variations in the health outcome explained, the number of admissions was tranformed by taking the logarithm and multiple regression was then applied. The value R2 was used to quantify the proportion of variations explained by the core model.
   
f) Effects were obtained per unit change in pollutant levels measured by current day and average of current day up to a previous day, called the best lag day, to be determined by Akaike掇 Information Criterion with the minimum value. The best lag day was determined up to previous 5 days for ozone and up to previous 3 days for the other pollutants.
   
g) Interaction effect between each pollutant (in the original scale) and each of the other pollutants was checked by first dichotomizing the other pollutants using the median and multiplying each of them to obtain the interaction term, and then introducing the interaction term into the model. When there was an interaction effect, the air pollutant effect was estimated for high and low levels of the other pollutant.
   
h) Interactions between pollutant concentrations and season were checked by defining a season dummy variable with 1 for cool (January-April; November-December) and 0 for warm (May-October) season, multiplying the season dummy variable with concentrations to obtain the interaction term and fit each interaction term to the model. When there was an interaction effect, relative risks in cool and warm seasons were obtained respectively.
   
i) Effects due to copollutants were estimated by putting the other pollutants one by one in the model. Effect with adjustment for a copollutant which was associated with the maximum adjusting effect was used.
   
j)

Autocorrelations were adjusted for using Newy-West method.6 Non-linear copollutant effects were studied by means of generalised additive model.7

Non-linear dose-response relationships of pollutants were also studied using generalised additive model.

   
k) Validation of the models for 1995-97 was achieved by comparing the observed numbers of health outcomes in the first half year of 1998 with the predicted from the model developed using 1995-97 data.

 

4.0 Findings

4.1

Descriptive statistics

Summary statistics of all the health outcomes used in this study are shown in Tables 5-6 below.

Table 5: Summary statistics of daily hospital admissions (including deaths) by disease groups (1995-97)

Admission No. (days) Mean SD Min P25 Median P75 Max
Respiratory 1096 203.7 38.49 116 175 197 227 335
Asthma 1096 24.7 8.59 7 19 23 29 63
COPD 1096 48.7 12.27 22 40 48 56 94
Cardiac 1096 98.7 23.28 40 80 101 114 179
IHD 1096 35.0 10.26 8 28 35 44 76
Death:
  • Cardiovascular
 1096 10.2 3.77 1 8 10 13 24
  • Respiratory
 1096 10.8 4.06 1 8 10 13 27

Min: Minimum; P25: 25% centile; P75: 75% centile; Max: Maximum
Note : COPD - Chronic obstructive pulmonary disease excluding asthma
IHD - Ischaemic heart disease

Table 6: Summary statistics of daily hospital admissions (including deaths) by disease groups (1998 Jan-June)

Admission No. (days) Mean SD Min P25 Median P75 Max
Respiratory 181 259.0 43.39 174 227 259 285 389
Asthma 181 23.7 7.2 10 19 22 28 48
COPD 181 54.1 11.86 32 45 53 62 90
Cardiac 181 104.1 22.27 48 88 106 118 166
IHD 181 38.3 10.52 15 31 39 45 65
Death:
  • Cardiovascular
 181 9.8 3.81 3 7 9 12 23
  • Respiratory
 181 11.6 4.15 3 8 11 14 23

Min: Minimum; P25: 25% centile; P75: 75% centile; Max: Maximum
Note : COPD - Chronic obstructive pulmonary disease excluding asthma
IHD - Ischaemic heart disease

Summary statistics of daily mean temperature, relative humidity, and mean concentration of each of the pollutants are shown in the Basic Tables B1-2.

The correlations among the five pollutants, and between pollutants and health outcomes are shown in Table 7 below.

Table 7: Spearman掇 rank correlation coefficient (r) between daily concentrations of pollutants, meteorological measures and hospital admissions (1995-97)

  Respiratory COPD* Asthma Cardiac IHD
  r (p-value) r (p-value) r (p-value) r (p-value) r (p-value)
SO2 (24-hr) 0.10 (0.00) 0.14 (0.00) -0.04 (0.15) 0.15 (0.00) 0.07 (0.02)
NO2 (24-hr) 0.11 (0.00) 0.24 (0.00) 0.27 (0.00) 0.23 (0.00) 0.15 (0.00)
RSP (24-hr) -0.01 (0.73) 0.13 (0.00) 0.27 (0.00) 0.16 (0.00) 0.11 (0.00)
O3 (8-hr) -0.07 (0.03) -0.08 (0.01) 0.26 (0.00) -0.02 (0.68) 0.01 (0.85)
FSP (24-hr) -0.06 (0.12) 0.10 (0.01) 0.17 (0.00) 0.08 (0.08) 0.02 (0.49)
Temperature -0.01 (0.76) -0.35 (0.00) -0.21 (0.00) -0.14 (0.00) -0.04 (0.16)
Humidity 0.10 (0.00) 0.04 (0.20) -0.30 (0.00) -0.07 (0.02) -0.06 (0.03)

* excluding asthma

Table 8: Matrix of Spearman掇 rank corelation coefficient (r) between daily concentrations of pollutants and meteorological measures (1995-97)

  SO2 RSP O3 (8-hr) FSP Temperature Humidity
NO2 0.37 0.82 0.43 0.79 -0.45 -0.35
SO2   0.30 -0.18 0.33 0.17 -0.16
RSP     0.54 0.91 -0.42 -0.53
O3 (8-hr)       0.47 -0.14 -0.59
FSP         -0.39 -0.39
Temperature           0.19

Note: all p values are <0.0001
   
4.2

Statistical modelling

4.2.1

Core models for hospital admissions

The core models, i.e. models with terms which explain long term trends, seasonal patterns and other variations due to potential confounding factors for daily counts of admission, are shown in Basic Tables E2-20. The numbers of cycles for all the 19 health outcomes are shown in Basic Table E1.

Overall, 8-9 numbers of cycles, each ranging from 0.33 to 5.98 cycles per year, were required to explain the seasonal patterns in each health outcome.

For all ages, the core models explained 41-72% of the total variations, in respiratory (R2 68%), COPD (53%), asthma (41%), cardiac (72%) and IHD (56%) admissons. In specific age groups, the core model explained less variations in the older than the younger age groups. In hospital deaths, the core model explained the least variations (26% for respiratory deaths; 21% for cardiovascular deaths).

Trends and seasonal patterns varied extensively among individual health outcomes.

Monday to Friday, except for asthma in all age groups, were associated with increased admissions relative to Sunday (0.000<p<0.799), and Saturday had different effects in different health outcomes. For asthma, Monday was associated with increased admissions (0.000<p<0.100), and Wednesday to Saturday were associated with decreased admissions relative to Sunday (0.000<p<0.939), while Tuesday did not show a clear pattern of association.

Holidays, except for asthma for all age groups and respiratory 0-14 age group, were associated with decreased admissions (0.000<p<0.045). Asthma all age groups and respiratory 0-14 age group were associated with increased admissions (0.002<p<0.288).

Days after holidays, except for IHD 65-74 age group (p=0.633), were associated with increased admissions (0.000<p<0.404).

Influenza, except for asthma 0-14 age group (p=0.098), was associated with increased admissions due to respiratory diseases (0.000<p<0.864).

Temperature, except for asthma all age groups, was associated with increased admissions (0.000<p<0.864). For asthma, it was associated with decreased admissions (0.000<p<0.134).

Humidity was associated with decreased admissions for all categories of respiratory diseases (0.000<p<0.689); and it did not show a consistent pattern of association with categories of cardiovascular diseases (p>0.198).
   
4.2.2

Core models for hospital deaths

Hospital deaths due to respiratory disease were associated with Monday, higher temperature and lower humidity (p<0.040); and deaths due to cardiovascular disease were associated with lower temperature and lower humidity (p<0.009).
   
4.2.3

Models for pollutants on hospital admissions

Effects of single pollutants at current day (lag 0) and average of best current day up to 3 or 5 previous days (best lag 0-3 or lag 0-5 days) are shown in Table 9.

Table 9: Relative risks (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase in a pollutant for each category of hospital admission ?1995-97 (Basic Table F1)

Respiratory disease (ICD9: 460-519)
    Lag 0     Lag (0-n)*  
Pollutant Age RR (95% CI) p-value Lag : RR (95% CI) p-value
NO2 (24-hr) All 1.08 (1.06,1.10) 0.0000 0-0 : 1.08 (1.06,1.10) 0.0000
  <15 1.04 (1.01,1.07) 0.0055 0-3 : 1.08 (1.04,1.12) 0.0001
  15-64 1.06 (1.03,1.10) 0.0002 0-0 : 1.06 (1.03,1.10) 0.0002
  65-74 1.07 (1.03,1.10) 0.0002 0-1 : 1.09 (1.06,1.15) 0.0000
  75+ 1.10 (1.07,1.14) 0.0000 0-0 : 1.10 (1.07,1.14) 0.0000
SO2 (24-hr) All 1.04 (1.02,1.07) 0.0024 0-0 : 1.04 (1.02,1.07) 0.0024
  <15 1.01 (0.96,1.05) 0.7884 0-1 : 0.99 (0.94,1.04) 0.6575
  15-64 1.04 (0.99,1.09) 0.0832 0-0 : 1.04 (0.99,1.09) 0.0832
  65-74 1.06 (1.01,1.11) 0.0239 0-0 : 1.06 (1.01,1.11) 0.0239
  75+ 1.14 (1.09,1.19) 0.0000 0-0 : 1.14 (1.09,1.19) 0.0000
RSP (24-hr) All 1.04 (1.02,1.06) 0.0000 0-3 : 1.06 (1.04,1.09) 0.0000
  <15 1.04 (1.01,1.07) 0.0019 0-3 : 1.09 (1.05,1.12) 0.0000
  15-64 1.03 (1.00,1.05) 0.0596 0-0 : 1.03 (1.00,1.05) 0.0596
  65-74 1.02 (0.99,1.05) 0.1248 0-1 : 1.05 (1.01,1.08) 0.0075
  75+ 1.05 (1.03,1.08) 0.0001 0-0 : 1.05 (1.03,1.08) 0.0001
O3 (8-hr) All 1.03 (1.01,1.05) 0.0005 0-2 : 1.08 (1.06,1.11) 0.0000
  <15 1.03 (1.00,1.06) 0.0326 0-2 : 1.07 (1.03,1.10) 0.0001
  15-64 1.00 (0.97,1.03) 0.9632 0-2 : 1.04 (1.00,1.07) 0.0413
  65-74 1.04 (1.01,1.07) 0.0182 0-2 : 1.09 (1.05,1.14) 0.0000
  75+ 1.00 (0.98,1.13) 0.7280 0-1 : 1.06 (1.06,1.10) 0.0010
FSP (24-hr) All 1.05 (1.02,1.07) 0.0002 0-3 : 1.06 (1.03,1.10) 0.0001
  <15 1.02 (0.98,1.06) 0.3335 0-3 : 1.06 (1.01,1.11) 0.0215
  15-64 1.02 (0.98,1.06) 0.2963 0-0 : 1.02 (0.98,1.06) 0.2963
  65-74 1.04 (0.99,1.09) 0.1026 0-1 : 1.07 (1.01,1.12) 0.0125
  75+ 1.08 (1.04,1.12) 0.0001 0-0 : 1.08 (1.04,1.12) 0.0001
COPD (ICD9: 490-496 excluding 493)
NO2 (24-hr) All 1.13 (1.09,1.16) 0.0000 0-1 : 1.14 (1.10,1.19) 0.0000
SO2 (24-hr) All 1.15 (1.10,1.20) 0.0000 0-0 : 1.15 (1.10,1.20) 0.0000
RSP (24-hr) All 1.05 (1.02,1.08) 0.0002 0-0 : 1.05 (1.02,1.08) 0.0002
O3 (8-hr) All 1.01 (0.98,1.03) 0.6988 0-3 : 1.08 (1.04,1.12) 0.0001
FSP (24-hr) All 1.06 (1.02,1.10) 0.0049 0-0 : 1.06 (1.02,1.10) 0.0049
Asthma (ICD9 493)
NO2 (24-hr) All 1.12 (1.07,1.17) 0.0000 0-3 : 1.22 (1.14,1.30) 0.0000
  <15 1.11 (1.04,1.19) 0.0014 0-3 : 1.33 (1.22,1.46) 0.0000
  15-64 1.12 (1.04,1.20) 0.0031 0-0 : 1.12 (1.04,1.20) 0.0031
SO2 (24-hr) All 1.04 (0.97,1.11) 0.2856 0-0 : 1.04 (0.97,1.11) 0.2856
  <15 0.96 (0.87,1.06) 0.3990 0-1 : 0.95 (0.84,1.06) 0.3547
  15-64 1.14 (1.03,1.27) 0.0116 0-0 : 1.14 (1.03,1.27) 0.0116
RSP(24-hr) All 1.05 (1.01,1.09) 0.0108 0-3 : 1.17 (1.11,1.23) 0.0000
  <15 1.09 (1.04,1.15) 0.0011 0-3 : 1.29 (1.21,1.38) 0.0000
  15-64 0.98 (0.92,1.04) 0.5091 0-0 : 0.98 (0.92,1.04) 0.5091
O3 (8-hr) All 1.04 (1.00,1.09) 0.0584 0-2 : 1.17 (1.11,1.23) 0.0000
  <15 1.10 (1.04,1.17) 0.0009 0-2 : 1.24 (1.16,1.34) 0.0000
  15-64 0.98 (0.92,1.05) 0.5653 0-2 : 1.09 (1.01,1.18) 0.0356
FSP (24-hr) All 1.00 (0.94,1.06) 0.9626 0-3 : 1.11 (1.03,1.20) 0.0066
  <15 1.03 (0.95,1.12) 0.4783 0-3 : 1.25 (1.13,1.38) 0.0000
  15-64 0.92 (0.84,1.01) 0.0804 0-0 : 0.92 (0.84,1.01) 0.0804
Cardiac disease (ICD9: 390-429)
NO2 (24-hr) All 1.06 (1.04,1.09) 0.0000 0-1 : 1.07 (1.04,1.10) 0.0000
  15-64 1.06 (1.02,1.09) 0.0029 0-0 : 1.06 (1.02,1.09) 0.0029
  65-74 1.05 (1.01,1.09) 0.0138 0-3 : 1.09 (1.03,1.15) 0.0018
  75+ 1.08 (1.05,1.12) 0.0000 0-0 : 1.08 (1.05,1.12) 0.0000
SO2 (24-hr) All 1.07 (1.03,1.10) 0.0001 0-0 : 1.07 (1.03,1.10) 0.0001
  15-64 1.03 (0.98,1.09) 0.1775 0-2 : 1.06 (0.99,1.14) 0.0712
  65-74 1.04 (0.98,1.09) 0.1714 0-1 : 1.06 (1.00,1.13) 0.0685
  75+ 1.09 (1.04,1.15) 0.0004 0-0 : 1.09 (1.04,1.15) 0.0004
RSP (24-hr) All 1.02 (1.01,1.05) 0.0137 0-1 : 1.03 (1.01,1.05) 0.0071
  15-64 1.02 (0.99,1.05) 0.1590 0-3 : 1.03 (0.99,1.07) 0.1481
  65-74 1.02 (0.99,1.05) 0.2054 0-1 : 1.03 (1.00,1.07) 0.0767
  75+ 1.04 (1.01,1.07) 0.0132 0-1 : 1.05 (1.01,1.08) 0.0000
O3 (8-hr) All 1.00 (0.98,1.02) 0.8749 0-5 : 1.04 (1.01,1.08) 0.0058
  15-64 1.01 (0.98,1.04) 0.4747 0-5 : 1.03 (0.98,1.08) 0.2297
  65-74 1.00 (0.97,1.04) 0.8842 0-5 : 1.08 (1.03,1.14) 0.0023
  75+ 0.99 (0.96,1.03) 0.7021 0-3 : 1.07 (1.03,1.11) 0.0016
FSP (24-hr) All 1.02 (0.99,1.05) 0.1452 0-3 : 1.05 (1.01,1.09) 0.0023
  15-64 1.01 (0.97,1.06) 0.6339 0-3 : 1.02 (0.96,1.08) 0.4812
  65-74 1.01 (0.96,1.06) 0.6678 0-3 : 1.03 (1.07,1.10) 0.2732
  75+ 1.05 (1.00,1.10) 0.0309 0-3 : 1.08 (1.03,1.14) 0.0044
IHD (ICD9: 410-414)
NO2 (24-hr) All 1.04 (1.00,1.07) 0.0552 0-0 : 1.04 (1.00,1.07) 0.0552
  15-64 0.99 (0.93,1.04) 0.6226 0-0 : 0.99 (0.93,1.04) 0.6226
  65-74 1.07 (1.01,1.13) 0.0273 0-3 : 1.11 (1.06,1.11) 0.0116
  75+ 1.07 (1.01,1.14) 0.0222 0-0 : 1.07 (1.01,1.14) 0.0222
SO2 (24-hr) All 1.01 (0.96,1.06) 0.6082 0-1 : 1.02 (0.96,1.08) 0.5635
  15-64 0.96 (0.89,1.04) 0.3536 0-0 : 0.96 (0.89,1.04) 0.3536
  65-74 1.09 (1.07,1.12) 0.0000 0-1 : 1.04 (0.95,1.14) 0.3964
  75+ 1.08 (1.05,1.10) 0.0000 0-3 : 0.93 (0.81,1.05) 0.2366
RSP (24-hr) All 1.03 (1.00,1.06) 0.0982 0-0 : 1.03 (1.00,1.06) 0.0982
  15-64 0.99 (0.95,1.04) 0.7484 0-1 : 0.98 (0.93,1.03) 0.4713
  65-74 1.05 (1.00,1.09) 0.0593 0-1 : 1.07 (1.03,1.12) 0.0161
  75+ 1.05 (1.00,1.10) 0.0478 0-0 : 1.05 (1.00,1.10) 0.0478
O3 (8-hr) All 1.00 (0.97,1.03) 0.9800 0-3 : 1.02 (0.98,1.07) 0.2944
  15-64 1.01 (0.96,1.06) 0.7614 0-1 : 1.01 (0.95,1.07) 0.7592
  65-74 1.01 (0.96,1.06) 0.8053 0-2 : 1.05 (0.99,1.12) 0.1246
  75+ 0.99 (0.94,1.04) 0.7276 0-3 : 1.04 (0.97,1.11) 0.3225
FSP (24-hr) All 0.99 (0.95,1.04) 0.6478 0-0 : 0.99 (0.95,1.04) 0.6478
  15-64 0.95 (0.89,1.02) 0.1785 0-0 : 0.95 (0.89,1.02) 0.1785
  65-74 1.00 (0.93,1.07) 0.9873 0-1 : 1.03 (0.95,1.11) 0.4381
  75+ 1.04 (0.96,1.12) 0.3375 0-1 : 1.05 (0.97,1.14) 0.2437

* Note:
  1. Lag 0-n denotes the cumulative effects of the mean pollutant concentration of lag from day 0 (same day) to day 1 up to day n
  2. using up to Lag 0-5 for ozone and up to Lag 0-3 for other pollutants

NO2 was associated with increased admissions for all health outcomes, except IHD 15-64 age group, at current day and best lag day (p<0.055).

SO2 was associated with increased admissions at the best lag day for respiratory disease all ages and the older than 15 age group (p<0.083), COPD all ages (p=0.000), asthma 15-64 age group (p=0.012) and cardiac disease all ages and all older than 15 age groups (p<0.071).

RSP was associated with increased admissions at the best lag day for respiratory disease in all ages and all the age groups (p<0.060), for COPD all ages (p=0.000), for asthma all ages and 0-14 age group (p=0.000), for cardiac disease in all ages and older than 65 age groups (p<0.077) and IHD all ages and older than 65 age groups (p<0.098).

O3 was not associated with admissions for most outcomes at the current day; but it was associated with most outcomes at the best lag days for respiratory diseases all ages and all age groups (p<0.041), COPD all ages (p=0.000), asthma all ages and all younger than 65 age groups (p<0.036) and cardiac disease all ages and all older than 65 age groups (p<0.006).

FSP was associated with admissions for most outcomes at the best lag days for respiratory diseases all ages, 0-14 and older than 65 age groups (p<0.022), for COPD (p=0.005), asthma all ages and all age groups (p<0.080) and for cardiac disease all ages and older than 75 age group (p<0.004).
   
4.2.4

Models for pollutants on hospital deaths

Hospital deaths due to respiratory disease all ages, were associated with NO2, SO2 and RSP (p<0.084). Hospital deaths due to cardiovascular disease all ages, were associated with NO2 and RSP (p<0.079). The relative risks with 95% confidence intervals are shown in Table 10 below

Table 10: Relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase in a pollutant for each category of hospital deaths 1995-97 (Basic Table F2) ?all ages

Respiratory disease (ICD9: 460-519)
  (Lag 0)     (Lag 0-n)*  
Pollutant RR (95% CI) p-value Lag : RR (95% CI) p-value
NO2 (24-hr) 1.08 (1.02,1.15) 0.0140 0-1 : 1.10 (1.03,1.18) 0.0068
SO2 (24-hr) 1.08 (1.00,1.18) 0.0638 0-0 : 1.08 (1.00,1.18) 0.0638
RSP (24-hr) 1.04 (0.99,1.09) 0.1655 0-1 : 1.05 (0.99,1.11) 0.0840
O3 (8-hr) 1.01 (0.95,1.06) 0.7951 0-3 : 1.06 (0.99,1.14) 0.1043
FSP (24-hr) 1.02 (0.94,1.10) 0.6398 0-1 : 1.04 (0.95,1.13) 0.3981
Cardiovascular disease (ICD9: 390-459)
NO2 (24-hr) 1.08 (1.02,1.15) 0.0130 0-0 : 1.08 (1.02,1.15) 0.0130
SO2 (24-hr) 1.01 (0.92,1.11) 0.8009 0-3 : 0.95 (0.83,1.09)  0.4560
RSP (24-hr) 1.02 (0.97,1.07) 0.4196 0-3 : 1.06 (0.99,1.14)  0.0789
O3 (8-hr) 1.00 (0.95,1.06) 0.9917 0-4 : 1.05 (0.97,1.14)  0.1914
FSP (24-hr) 1.02 (0.94,1.11) 0.6596 0-3 : 1.05 (0.95,1.17) 0.3150

* Note:
  1. Lag 0-n denotes the cumulative effects of the mean pollutant concentration of lag from day 0 (same day) to day 1 up to day n
  2. using up to Lag 0-5 for ozone and up to Lag 0-3 for other pollutants
   
4.2.5

Effects of  pollutants adjusted for copollutants

Effect of pollutant with adjustment for each of the other pollutants at their best lag days are shown in Tables 11-12 below. Only hospital admissions and deaths for all ages are presented.

Table 11: Relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase in a pollutant adjusted for a copollutant which was associated with the greatest adjustment effect ?hospital admissions 1995-97

  Pollutant RR (95% CI) p-value [Copollutant]
Respiratory disease all ages
  NO2 1.06 (1.04, 1.08) 0.000 [O3]
  SO2 0.96 (0.92, 0.99) 0.016 [NO2]
  RSP 1.03 (1.00, 1.06) 0.023 [O3]
  O3 1.07 (1.04, 1.09) 0.000 [NO2]
  FSP 1.04 (0.98, 1.10) 0.192 [RSP]
COPD all ages
  NO2 1.11 (1.06, 1.16) 0.000 [SO2]
  SO2 1.08 (1.02, 1.13) 0.005 [NO2]
  RSP 0.98 (0.95, 1.01) 0.253 [NO2]
  O3 1.03 (0.99, 1.07) 0.123 [NO2]
  FSP 1.02 (0.95, 1.09) 0.655 [RSP]
Asthma all ages
  NO2 1.13 (1.02, 1.25) 0.018 [RSP]
  SO2 0.96 (0.90, 1.04) 0.316 [NO2]
  RSP 1.09 (1.00, 1.18) 0.045 [NO2]
  O3 1.11 (1.05, 1.18) 0.000 [RSP]
  FSP 0.99 (0.90, 1.10) 0.913 [NO2]
Cardiac disease all ages
  NO2 1.06 (1.02, 1.09) 0.001 [SO2]
  SO2 1.03 (0.99, 1.07) 0.128 [NO2]
  RSP 0.98 (0.95, 1.01) 0.244 [NO2]
  O3 1.02 (0.99, 1.05) 0.213 [NO2]
  FSP 1.01 (0.96, 1.05) 0.770 [NO2]
IHD all ages
  NO2 1.03 (0.98, 1.08) 0.289 [RSP]
  SO2 0.99 (0.93, 1.06) 0.858 [RSP]
  RSP 1.01 (0.97, 1.05) 0.664 [NO2]
  O3 1.01 (0.97, 1.06) 0.603 [NO2]
  FSP 1.00 (0.95, 1.06) 0.905 [SO2]

NO2 was associated with all categories of respiratory admissions including COPD and asthma (p<0.018) and with cardiac admissions (p=0.001).

SO2 was associated with admissions due to respiratory disease (p=0.016) and COPD (p=0.005).

RSP was associated with admissions due to respiratory disease (p=0.023) and asthma (p=0.045).

O3 was associated with admissions due to respiratory disease (p=0.000) and asthma (p=0.000).

FSP was not associated with any of the admission categories under study (p>0.192).

No pollutants, except NO2, were associated with admissions due to cardiac disease (p>0.128); and no pollutants were associated with admissions due to IHD (p>0.289).

Table 12: Relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase in a pollutant adjusted for a copollutant which was associated with the greatest adjustment effect ?hospital deaths 1995-97

  Pollutant RR (95% CI) p-value [Copollutant]
Respiratory deaths all ages
  NO2 1.09 (1.00, 1.19) 0.044 [SO2]
  SO2 1.02 (0.92, 1.13) 0.669 [NO2]
  RSP 0.99 (0.91, 1.07) 0.761 [NO2]
  O3 1.03 (0.95, 1.11) 0.457 [NO2]
  FSP 1.00 (0.90, 1.12) 0.959 [O3]
Cardiovascular deaths all ages
  NO2 1.07 (1.00, 1.15) 0.057 [RSP]
  SO2 0.87 (0.74, 1.02) 0.079 [RSP]
  RSP 1.03 (0.95, 1.11) 0.449 [NO2]
  O3 1.03 (0.94, 1.12) 0.564 [RSP]
  FSP 1.00 (0.89, 1.13) 0.967 [RSP]

NO2 was associated with respiratory deaths (p=0.044) and with cardiovascular deaths (p=0.057) for all ages. For the other pollutants they were not associated with both respiratory and cardiovascular deaths (p>0.079).

   
4.2.6

Models with adjustment for autocorrelation

Figures 5 (a) - (e) show the residual plots and partial autocorrelation function plots for all the hospital admission categories. There were some but non-regular patterns in the residual plots and some autocorrelation (r<0.4) in respiratory disease for all ages and all age groups and in asthma 0-14 age group. For the other admission categories, there were no patterns in the residual plots with very little autocorrelation (r<0.25).

Adjustment for autocorrelation did not affect the RR estimates; but it would affect the standard errors, if autocorrelation exists, and therefore the 95% confidence intervals will be wider and the p-values larger. The estimated 95% CIs and p-values, if they differred from the original estimates after the adjustment, are shown in Basic Tables G1.

Hospital deaths did not show autocorrelations in the residuals.
   
4.2.7

Non-linear effects

Generalized additive modelling (GAM) was used to assess the sensitivity of using non-linear assumption for meteorological covariates (temperature and humidity) on effect estimates of pollutants. The RRs were practically the same as those presented in Tables 11-12. Those which show some slight variations were COPD due to O3 (1.04 in GAM estimate vs 1.03 in original estimate); asthma due to NO2 (1.10 vs 1.13), SO2 (0.97 vs 0.96) and O3 (1.10 vs 1.11); cardiac disease due to O3 (1.03 vs 1.02) and IHD due to SO2 (1.00 vs 0.99). (Data were not presented in table.)

Non-linear dose-response relationships for all ages and in models with adjustment for copollutant were found in NO2 on asthma and respiratory deaths (p<0.008); RSP on respiratory and asthma admissions (p<0.036); and FSP on respiratory admissions (p=0.040). The others were non-significant (p>0.050). (Data were not presented in table.)
   
4.2.8

Models for interactions between pollutants

Only admissions for all ages were studied and those significant interactions (p<0.05) between pollutants were presented in Table 13 below.

Table 13: Relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase of a pollutant at low and high level of a copollutant ?hospital admission 1995-97 (Basic Table H1)

1. Respiratory all ages

Copollutant NO2 RR (95% CI) p-value
SO2 Low 1.12 (1.08, 1.16) 0.000
  High 1.07 (1.05, 1.10) 0.000

2. COPD all ages

Copollutant NO2 RR (95% CI) p-value
O3 Low 1.09 (1.05, 1.14) 0.000
  High 1.17 (1.12, 1.23) 0.000
    SO2 RR (95% CI)  
NO2 Low 1.22 (1.12, 1.33) 0.000
  High 1.08 (1.03, 1.14) 0.002
    O3 RR (95% CI)  
NO2 Low 0.94 (0.90, 0.99) 0.011
  High 1.01 (0.97, 1.04) 0.739

3. Asthma all ages

Copollutant NO2 RR (95% CI) p-value
SO2 Low 1.27 (1.16, 1.39) 0.000
  High 1.10 (1.04, 1.17) 0.002
    O3 RR (95% CI)  
SO2 Low 1.10 (1.04, 1.16) 0.001
  High 1.00 (0.95, 1.06) 0.893

Interactions between pollutants were found in admissions due to the respiratory disease (between SO2 ?NO2); due to COPD (O3 ?NO2; NO2 ?SO2; NO2 ?O3) and due to asthma (SO2 ?NO2; SO2 ?O3), but they were not found in admissions due to cardiovascular disease.

 
   
4.2.9

Models with interaction between pollutants and seasons

Only admissions for all ages were studied and those with a significant difference (p<0.05) between cool and warm seasons are presented in Table 14.

Table 14: Relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase of a pollutant in cool and warm seasons ?hospital admission 1995-97 (Basic Table I1)

a. Respiratory all ages

Relative Risk (95% CI)
Pollutant Cool Warm
NO2 1.10 (1.07, 1.13) 1.05 (1.02, 1.08)
FSP 1.09 (1.04, 1.53) 0.99 (0.93, 1.05)

a. COPD all ages

Relative Risk (95% CI)
Pollutant Cool Warm
NO2 1.17 (1.12, 1.22) 1.07 (1.02, 1.12)
O3 1.07 (1.04, 1.11) 1.02 (0.97, 1.06)

b. Cardiac all ages

Relative Risk (95% CI)
Pollutant Cool Warm
NO2 1.11 (1.07, 1.14) 1.01 (0.98, 1.04)
SO2 1.11 (1.06, 1.16) 1.04 (0.99, 1.08)
RSP 1.03 (1.00, 1.06) 0.98 (0.95, 1.00)
O3 1.05 (1.02, 1.08) 0.99 (0.96, 1.02)

b. IHD all ages

Relative Risk (95% CI)
Pollutant Cool Warm
NO2 1.08 (1.03, 1.14) 0.99 (0.94, 1.04)
O3 1.05 (1.01, 1.09) 0.99 (0.95, 1.04)

When there were interactions, the effects were stronger and were greater than unity in the cool season than in the warm season.
   
4.3

Validity and reliability of models

4.3.1

Predictive validity of 1995-97 models

Figures 7 (a)-(e) depict the discrepancies between the observed in the first half year of 1998 and the predicted counts of hospital admissions due to the 5 categories of diseases for all ages, according to the model derived from 1995-97 data. Figures 8(a)-(b) depict the discrepancies for hospital deaths due to respiratory and cardiovascular diseases.

For admissions due to respiratory disease, the model (core model plus a pollutant) predicted the admission counts closely for the first quarter of the year; but under-estimated the numbers markedly during the fourth and fifth months of 1998 (Figure 7(a)).

For COPD, the 1995-97 models predicted the admission counts closely (Figure 7(b)).

For asthma, the 1995-97 models predicted the admission counts closely except in the fourth month of 1998 (Figure 7(c)).

For cardiac, the 1995-97 models predicted the admission counts well with a small amount of the observed above the predicted. The discrepancies were consistent throughout the first half year period of 1998 (Figure 7(d)).

For IHD, the prediction was similar to that for cardiac admissions (Figure 7(e)).

For deaths due to respiratory disease, the 1995-97 models under estimated the observed especially during the second month of 1998 (Figure 8(a)).

For deaths due to cardiovascular disease, the 1995-97 models predicted the observed counts closely (Figure 8(b)).

Overall, all the models appear to predict the observed closely except those for admissions and deaths due to respiratory disease.

   
4.3.2

Comparison of 1995-97 with 1995-96 models

Estimates for effects of each pollutant for health outcomes defined as in the previous study, at the best lag day, are compared between those estimated from the 1995-96 and 1995-97 models in Tables 15-16.

Table 15: Comparison between relative risks (RR) and 95% confidence intervals (95% CI) for 50 ug/m3 increase in a pollutant estimated from the 1995-96 and 1995-97 models ?hospital admissions

(i) Cardiovascular + Respiratory
  995-96 (Lag 0-n)#1 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-0 :1.07 (1.04,1.09)*** 0-0 :1.06 (1.05,1.08)***
SO2 (24-hr) 0-0 :1.04 (1.01,1.07)* 0-0 :1.04 (1.02,1.07)***
RSP (24-hr) 0-1 :1.04 (1.02,1.06)*** 0-3 :1.05 (1.03,1.07)***
O3 (8-hr) 0-2 :1.07 (1.05,1.10)*** 0-3 :1.05 (1.03,1.07)***
(ii) Cardiovascular
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-2 :1.08 (1.04,1.11)*** 0-0 :1.04 (1.02,1.06)***
SO2 (24-hr) 0-1 :1.05 (1.01,1.10)* 0-0 :1.05 (1.02,1.08)***
RSP (24-hr) 0-3 :1.03 (1.00,1.06)* 0-3 :1.03 (1.00,1.05)*
O3 (8-hr) 0-5 :1.07 (1.03,1.10)*** 0-5 :1.03 (1.00,1.05)*
(iii) Respiratory
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-0 :1.08 (1.05,1.11)*** 0-0 :1.08 (1.06,1.10)***
SO2 (24-hr) 0-0 :1.03 (1.00,1.08) 0-0 :1.04 (1.02,1.07)**
RSP (24-hr) 0-1 :1.05 (1.02,1.07)*** 0-3 :1.06 (1.04,1.09)***
O3 (8-hr) 0-2 :1.10 (1.07,1.13)*** 0-2 :1.08 (1.06,1.11)***
(iv) Asthma
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag RR (95% CI)
NO2 (24-hr) 0-3 :1.12 (1.02,1.22)* 0-3 :1.22 (1.14,1.30)***
SO2 (24-hr) 0-3 :0.91 (0.78,1.06) 0-0 :1.04 (0.97,1.11)
RSP (24-hr) 0-3 :1.10 (1.03,1.18)** 0-3 :1.17 (1.11,1.23)***
O3 (8-hr) 0-2 :1.16 (1.08,1.24)*** 0-2 :1.11 (1.03,1.20)**
(v) Chronic obstructive pulmonary disease
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-0 :1.13 (1.09,1.17)*** 0-0 :1.11 (1.08,1.14)***
SO2 (24-hr) 0-0 :1.12 (1.06,1.18)*** 0-0 :1.10 (1.06,1.14)***
RSP (24-hr) 0-3 :1.08 (1.04,1.13)*** 0-3 :1.09 (1.06,1.13)***
O3 (8-hr) 0-2 :1.14 (1.09,1.18)*** 0-3 :1.10 (1.06,1.13)***
(vi) Ischaemic heart disease
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-2 :1.09 (1.03,1.16)** 0-0 :1.04 (1.00,1.07)
SO2 (24-hr) 0-1 :1.03 (0.95,1.12) 0-1 :1.02 (0.96,1.08)
RSP (24-hr) 0-1 :1.04 (1.00,1.09) 0-3 :1.02 (0.98,1.07)
O3 (8-hr) 0-2 :1.04 (0.99,1.10) 0-0 :0.99 (0.95,1.04)

* p < 0.05; ** p < 0.01; *** p < 0.001

# Note:
  1. Lag 0-n denotes the cumulative effect of the mean pollutant concentration of lag from day 0 (same day) up to previous n days
  2. the cumulative lag days were up to 5 days for ozone and up to 3 days for other pollutants

Overall, the effect estimates from the two sets of models were comparable with their corresponding 95% confidence intervals all overlapping the other.

Table 16: Comparison between relative risk (RR) and 95% confidence interval (95% CI) for 50 ug/m3 increase in a pollutant estimated from 1995-96 and 1995-97 models ?hospital deaths

(i) Cardiovascular + Respiratory
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-1 :1.12 (1.05,1.20)*** 0-1 :1.09 (1.04,1.15)***
SO2 (24-hr) 0-3 :0.93 (0.81,1.06) 0-0 :1.05 (0.99,1.12)
RSP(24-hr) 0-1 :1.04 (0.99,1.09) 0-3 :1.06 (1.01,1.11)*
O3 (8-hr) 0-2 :1.07 (1.05,1.10)*** 0-4 :1.05 (1.00,1.11)
(ii) Cardiovascular
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-1 :1.10 (1.01,1.21)* 0-0 :1.08 (1.02,1.15)*
SO2 (24-hr) 0-3 :0.89 (0.74,1.07) 0-3 :0.95 (0.83,1.09)
RSP (24-hr) 0-3 :1.05 (0.96,1.14) 0-3 :1.06 (0.99,1.14)
O3 (8-hr) 0-5 :1.15 (1.03,1.28)** 0-4 :1.05 (0.97,1.14)
(iii) Respiratory
  1995-96 (Lag 0-n)# 1995-97 (Lag 0-n)#
Pollutant Lag :RR (95% CI) Lag :RR (95% CI)
NO2 (24-hr) 0-1 :1.14 (1.04,1.25)** 0-1 :1.10 (1.03,1.18)**
SO2 (24-hr) 0-0 :1.08 (0.96,1.22) 0-0 :1.08 (1.00,1.18)
RSP (24-hr) 0-1 :1.05 (0.97,1.13) 0-1 :1.05 (0.99,1.11)
O3 (8-hr) 0-5 :1.22 (1.09,1.35)*** 0-3 :1.06 (0.99,1.14)

* p < 0.05;
** p < 0.01;
*** p < 0.001

# Note:
  1. Lag 0-n denotes the cumulative effect for the mean pollutant concentration of lag from day 0 (same day) up to previous n days
  2. the cumulative lag days were up to 5 days for ozone and up to 3 days for other pollutants

Overall, for deaths the effect estimates derived from both 1995-96 and 1995-97 models were comparable with the corresponding 95% confidence intervals overlapping each other. However the RRs for ozone decreased substantially in the 1995-97 estimates.

5.0 Discussion

5.1

Validity and reliability of the models

Predictive validity: The models derived from the 1995-97 data showed a high level of predictive validity (much better that those derived from 1995-96 data) although there are some slight distinguishable discrepancies between the observed and predicted for some categories of health outcomes.

The discrepancies may be due to incomplete return of the discharge diagnoses for 1998 validation data at the time of data collection. These may also be due to inadequacy of the models for some health outcomes (3 years of data are minimum for development of a predictive model). The discrepancies may reflect some changes which might have occurred in the environment. In any case, the presence of these discrepancies would not invalidate the models in estimating the health effects of air pollution.

Reliability between years of the study: The relative risks estimated from 1995-96 and from 1995-97 models are highly reproducible. In most cases, th RRs were within 5% of each other in the two sets of estimates; but for asthma the discrepancies ranged from 5% to 13%. These discrepancies may just be random.

   
5.2

Comparison with other APHEA studies

Respiratory admissions: The RRs (95% CI) summarized from four to five European cities8 as compared to those from this Hong Kong study are shown in Table 17.

Table 17: Relative risk (RR) with 95% confidence interval (95% CI) for 50 ug/m3 increase of a pollutant on daily respiratory admissions summarized from four to five European cities compared to those from this Hong Kong study

    European   Hong Kong
    RR (95% CI)   RR (95% CI)
NO2 15-64 1.00 (0.99,1.11)   1.06 (1.03,1.10)
  65+ 1.02 (0.98,1.06)   1.11 (1.08,1.15)
SO2 15-64 1.01 (0.99,1.03)   1.04 (0.99,1.09)
  65+ 1.02 (1.01,1.05)   1.11 (1.07,1.15)
TSP 15-64 1.01 (0.99,1.03) RSP 1.03 (1.00,1.05)
  65+ 1.02 (0.99,1.04)   1.06 (1.03,1.08)
O3 15-64 1.03 (1.01,1.05)   1.04 (1.00,1.07)
  65+ 1.04 (1.02,1.06)   1.07 (1.04,1.11)

Both European and Hong Kong studies demonstrated that air pollutants had a stronger effect in the 65+ than in the 15-64 age group. The effects in Hong Kong were consistently stronger than those in European cities.

Asthmatic admissions: The RRs (95% CI) summarized from four European cities,9 compared to those from this Hong Kong study, are shown in Table 18.

Table 18: Relative risk (RR) with 95% confidence interval (95% CI) for 50 ug/m3 increase of a pollutant on daily asthmatic admissions summarized from four European cities compared to those from this Hong Kong study

    European   Hong Kong
    RR (95% CI)   RR (95% CI)
NO2 <15 1.04 (1.00,1.07)   1.33 (1.22,1.46)
  15-64 1.04 (1.01,1.07)   1.12 (1.04,1.20)
SO2 <15 1.06 (1.00,1.13)   0.95 (0.84,1.06)
  15-64 1.00 (0.96,1.05)   1.14 (1.03,1.27)
BS <15 1.05 (0.98,1.12) RSP 1.29 (1.21,1.38)
  15-64 1.03 (0.98,1.08)   0.98 (0.92,1.04)
O3 <15 1.01 (0.98,1.04)   1.24 (1.16,1.34)
  15-64 1.04 (0.94,1.14)   1.09 (1.01,1.18)

For the 0-14 age group, the effect estimates due to NO2, RSP (compared with black smoke) and O3 were significantly greater in the Hong Kong than in the European studies. For the 15-64 age group the effects due to NO2, SO2 and O3 were also greater in Hong Kong than in European cities but thee differences were not statistically significant. In the Hong Kong study the differences between the <15 and 15-64 age groups were great.

COPD admissions: The RRs (95% CI) summarized from four to six European cities,10 compared to those from this Hong Kong study, are shown in Table 19.

Table 19: Relative risk (RR) with 95% confidence interval (95% CI) for 50 ug/m3 increase of a pollutant on daily COPD admissions summarized from four to six European cities compared to those from this Hong Kong study

    European   Hong Kong
    RR (95% CI)   RR (95% CI)
NO2 All ages 1.03 (1.00,1.04) NO2 1.14 (1.10,1.19)
SO2 All ages 1.02 (1.00,1.05) SO2 1.15 (1.10,1.20)
TSP All ages 1.03 (0.99,1.07) RSP 1.05 (1.02,1.08)
O3 All ages 1.05 (1.02,1.08) O3 1.08 (1.04,1.12)

The effects of pollutants estimated from the Hong Kong study were all greater than those estimated from the European studies. The differences were significant for NO2 and SO2.

Cardiovascular admissions: There are no equivalent estimates from APHEA studies published yet. From an American study11 the RR (95% CI) for cardiac admissions due to RSP was 1.06 (1.01,1.12) in 65+ age group as compared to that estimated from this Hong Kong study of 1.03 (1.00,1.07) in 65-74 and 1.05 (1.01,1.08) in 75+ age groups.
   
5.3

Limitations of the Hong Kong study

Non-linear effects: In the main analysis of this study, a linear association between air pollutants and health risks was assumed. However attempts were made to assess whether there were any non-linear dose-response relationships and to assess the sensitivity of using a non-linear assumption for the effect of meteorological variables, using generalized additive modelling. In this study the dose-response relationships between pollutants and health outcomes were reasonably linear and estimates were robust to non-linearity in the temperature and humidity covariates used in the models.

Unexplained patterns: Attempts were made in this study to estimate the numbers of cycles present in the time-series data in all core models systematically. However, there were still some discernible patterns in the residuals of the core models for respiratory admissions, indicating that the amount of data were not sufficient to establish the models in these outcomes. Time-series data longer than three years are required. But it is not advisable to eliminate all discernible patterns by using arbitrarily defined dummy variables, for this would lead to problems of over fitting and will disable the models to be used for prediction purposes.

Effects on mortality: Only deaths after admission to acute hospitals (accounting for 50% of all deaths in Hong Kong) were analysed in this study. As daily mortality data is now available from the Census and Statistics Department, they should be studied in order to be in line with the APHEA protocol.

   
5.4

Significance of study results

The effects of air pollutants estimated from this study were to a large extent greater than those summarised from APHEA studies in Europe. It is difficult to assess the reasons which may contribute to the greater effects in Hong Kong. The general explanations might include over-crowding in the population, easy access to and high utilization rates of hospital services and the Government announcement of the air pollution index which raised the awareness of the public to air pollution. Further detailed analysis and interpretation of the possible public health significance of this observation were beyond the scope of this study.

There are no studies from the Asia Pacific region, except one similar study from Sydney Australia12, for us to make comparison with. In the Sydney study, NO2 has a primary effect on hospital admissions for asthma, has a stronger effect than particles on COPD admissions, and also has effects on heart disease admissions (ICD9 410, 413, 427 and 428). The results from the Sydney study were similar to those from this Hong Kong study. In Hong Kong NO2 had an effect on asthma, COPD and cardiac disease (ICD9 390-429) (including the heart disease categories studied in Sydney), but in addition it was also associated with admissions for a broad respiratory disease category. Other pollutants including SO2, RSP and O3 were also associated with some disease categories. They could be considered as proxy indices for the effects of the complex mix of air pollution in large urban centre like Sydney.12

6.0 Conclusions

The results from this study show that short-term effects of air pollutants on health are strong. Given the consistency and reproducibility of the associations between air pollutants and cardio-respiratory diseases and the biological plausibility, these short term effects are likely to be causal. It is both feasible and necessary to develop further studies on the health effects of air pollution with a longer series of data, using the methodology adopted by the APHEA protocol, to strengthen and clarify some of the relationships observed in this study. Territory-wide SAR mortality should also be included as one of the main health outcomes to be studied. This would provide a sound basis on which to assess the health gains from future air quality interventions in Hong Kong.

References

  1. Wong CM, Ma S, Hedley AJ, Lam TH. Short-term effects of ambient air pollution on public health in Hong Kong ?a follow-up study. A consultancy report submitted to the Environmental Protection Department, Hong Kong Government, 1998, 29pp with Figures, Tables and Appendices.
     
  2. Wong TW, Ho KM, Lau TS, Neller A, Wong SL, ITS Yu. A study of short-term effects of ambient air pollution on public health. Report submitted to Environmental Protection Department, Hong Kong Government, 1997, 78pp.
     
  3. APHEA2. Short-Term Effects of Air Pollution on Health: A European Approach to Methodology, Dose-Response Assessment and Evaluation of Public Health Significance, 1998.
     
  4. McCullagh P, Nelder JA. Generalized Linear Models (2nd Edition). Cambridge: Chapman and Hall, 1989.
     
  5. Diggle PJ. Time Series: A Biostatistical Introduction. New York: Oxford University Press, 1990.
     
  6. Newy WK, West KD. A simple positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. Econometrica 1987; 55:703-708.
     
  7. Hastie TJ, Tibshirani RJ. Generalized Additive Models. London: Chapman and Hall, 1990.
     
  8. Spix C, Anderson HR, Schwartz J, Vigotti MA, LeTertre A, Vonk JM. Short-term effects of air pollution on hospital admissions of respiratory diseases in Europe: a quantitative summary of APHEA study results. Archives of Environmental Health 1998; 53:54-64.
     
  9. Sunyer J, Spix C, Quénel, Ponce-deLeón A, Pönka A, Barumandzadeh T, Gouloumi G, Bacharova L, Wojtyniak B, Vonk J, Bisanti L, Schwartz J, Katsouyanni K. Urban air pollution and emergency admissions for asthma in four European cities: the APHEA project. Thorax 1997; 52:760-765.
     
  10. Anderson HR, Spix C, Medina S, Schouten JP, Castellsague J, Rossi G, Zmirou D, Touloumi G, Wojtyniak B, Pönka A, Bacharova L, Schwartz J, Katsouyanni K. Air pollution and daily admissions for chronic obstructive pulmonary disease in 6 European cities: results from the APHEA project. Eur Respir J 1997; 10:1064-1071.
     
  11. Schwartz J. Air pollutant and hospital admissions for cardiovascular disease in Tucson. Epidemiology 1997; 8:371-377.
     
  12. Morgan G, Corbett S, Wlodarczyk J. Air pollution and hospital admission in Sydney, Australia, 1990 to 1994. American Journal of Public Health 1998; 88:1761-1766.

 
Back to topdot_clear.gifBackTable of ContentNext