Cardiovascular Events During General Elections In Bayamon, Puerto Rico

Abstract

Emotional stress has been linked to acute coronary events. We examined whether the emotional response to elections in Puerto Rico induced a similar response.

Methods

We reviewed records at HIMA San Pablo Hospital (HIMASP) and Ramon Ruiz Arnau University Hospital (HURRA) in Bayamón and identified patients admitted with ICD-9 codes 410, 411, and 413 or corresponding diagnoses during a period surrounding the general elections and compared them with the same time period in non-election years.

Results

Cardiovascular events accounted for 3.24% of election-year admissions vs. 5.51% during non-election years in HURRA (p=0.036, N=37), while accounting for 2.86% of election-year admissions in HIMASP vs. 3.27% during non-election years (non-significant).

Discussion

There was a trend towards a lower rate of admission for cardiovascular events during general elections in both hospitals, reaching statistical significance at HURRA. Further study may elucidate reasons for this behavior and determine whether similar trends hold true in other populations.

INTRODUCTION

The association between emotional and cardiovascular events has long been suspected and proposed. There are studies linking anger, hostility, and depression to increased rates of coronary heart disease [1, 2]. Several studies have also documented a link between typically stressful situations and coronary events. In 2005 Jette Moller and colleagues reported that sudden increases in work demands were associated with six times the risk of myocardial infarction [3]. Ching-Hong Tsai and colleagues reported in 2004 that admissions for myocardial infarction rose significantly following the Taiwan earthquake on September 21, 1999 [4], and similar findings were reported when coronary events were studied following the Iraqi missile war [5]. While these studies all support the role stress as a trigger for coronary events, they fundamentally represent a physiologic response to unexpected events. One might consider whether anticipated events in which subjects participate willingly would provoke similar responses.

Several studies evaluating coronary events surrounding football matches in Europe shed light this subject. One study performed in the German population showed significant increases in the incidence of myocardial infarction and arrhythmia in the days surrounding the 2006 World Cup [6]. Similar studies carried out in English and Dutch populations showed similar trends [7,8]. In this paper we present data whether general elections in Puerto Rico, events which are highly charged with emotion and in which subjects also participate willingly, are associated with an increase in acute coronary events.

METHODS

We were interested in examining different sectors of our community and as such we selected two tertiary hospitals in the region of Bayamon, Puerto Rico for our study. The first was the University Hospital Ramón Ruiz Arnau (HURRA). This is a teaching hospital which is sponsored and funded by the Department of Health of Puerto Rico. The second hospital was HIMA San Pablo in Bayamon (HIMASP). This is private hospital which is located within 3 miles of the first. Both receive and treat adult patients with acute cardiovascular events. Utilizing the International Classifications of Diseases sourcebook, we identified the ICD-9 codes which are commonly used to codify admitting and discharge diagnoses for acute cardiovascular events in the hospitals of Puerto Rico. From this sourcebook we selected the following codes which were felt to identify patients with acute coronary events.

410: Acute Myocardial Infarction

411: Acute Subacute Ischemic Heart Disease

413: Angina Pectoris

We selected a time period corresponding to seven days before and after the general elections in Puerto Rico for the years 2000 and 2004 in both hospitals, and we included the 2008 year in the data abstracted from HURRA. We tabulated the total number of adult admissions corresponding to the selected ICD-9 codes or any of their subcategories in both hospitals during the specified time periods. As a control group we studied the admissions with identical ICD-9 codes during a similar period of time for the non election years of 1999, 2001, 2003, 2005 and 2007. The specific dates in question were as follows: October 31, 1999 – November 14, 1999; October 31, 2000 – November 14, 2000 (General elections held November 7, 2000); October 31, 2001 – November 14, 2001; October 26, 2003 – November 9, 2003; October 26, 2004 – November 9, 2004 (General elections held November 2, 2004); October 26, 2005 – November 9, 2005; October 29, 2007 – November 11, 2007; October 29, 2008 – November 11, 2008 (General elections held November 4, 2008).

The patients were identified and selected from a perusal of the electronic data banks in HIMASP and through a manual review of admission and discharge records in HURRA. When available, the age and sex of the patients was recorded and tabulated. Only patients over 18 years of age were used in the study. We also identified the total number of adult admissions during the specified time frame in both hospitals in order to analyze proportions with acute coronary events. There was some doubt on the exact number of adult admissions between 2000 and 2004 for the time periods studied in HIMASP. This was related to the fact that the raw number of total admissions in these years was a total raw number that did not exclude patients under the age of 18. We decided to analyze the number of admissions during the months of November and December between 2005-2008, determine the proportion of adults over the age 18 admitted during this time frame, take an average proportion of admissions over the age 18 vs. under the age of 18 and use that number to estimate the number of adult admissions during the index elections years.

Statistical Methods

To describe the study group, percentages and descriptive analysis were used. In order to select the correct parametric or non parametric test, a normality test was performed in all the variables involved in our study. Differences among medians were evaluated using the two-independent samples Mann-Whitney U test. The significance level was set in 0.05. The Statistical Package for the Social Sciences (SPSS 14, Chicago IL) was used to analyze the data.

RESULTS

UNIVERSITY HOSPITAL RAMON RUIZ ARNAU (HU-RRA), See TABLE I

Table I.

Patients Admitted to HURRA with a Diagnosis of Ischemic Heart Disease. The numbers in parentheses represent the percentage of total adult admissions for that particular diagnosis and year. The numbers outside parentheses represent the actual number of patients admitted.

Dx	Year	1999	2000	2001	2003	2004	2005	2007	2008
410		8	1	3	0	2	4	1	1
		(3.94%)	(0.60%)	(1.81%)	(0.0%)	(2.56%)	(7.27%)	(1.64%)	(2.94%)
411		0	0	0	0	0	1	2	0
		(0.0%)	(0.0%)	(0.0%)	(0.0%)	(0.0%)	(1.82%)	(3.28%)	(0.0%)
413		4	5	4	1	0	0	0	0
		(1.97%)	(3.01%)	(2.41%)	(4.35%)	(0.0%)	(0.0%)	(0.0%)	(0.0%)
Total		12	6	7	1	2	5	3	1
		(5.91%)	(3.61%)	(4.22%)	(4.35%)	(2.56%)	(9.09%)	(4.92%)	(2.94%)
Admissions		203	166	166	23	78	55	61	34

During election years a total of 9 out of 278 admissions with acute coronary events were documented. This number accounts for 3.24% of adult admissions during this time. There were 6 men and 3 women with a mean patient age of 55 years. During non-election years there were 28 out of 508 adult admissions accounting for 5.51% of all adult admissions related to acute coronary events. Of these, 18 (64%) were male and 10 (36%) were female. Mean patient age was 66 (SD, 13). Thus during election years the number admissions due to acute coronary events was 3.24% as compared to 5.51% in the control years (See Figure 1). The Mann-Whitney U test analysis comparing the rate of election year admissions for cardiovascular events to those on non-election years showed significant difference. (p=0.036).

Figure 1.

Comparison of Admission Rates for Cardiovascular Events during Election Years and Non-Election Years at HIMASP and HU-RRA. (HIMASP=HIMA San Pablo Hospital. HURRA=Hospital Universitario Ramon Ruiz Arnau. NS=Non-significant.)

AUTHOR: Arnaldo E. Pérez-Mercado, M.D.

HIMA SAN PABLO BAYAMON (HIMASP), See TABLE II

Table II.

Patient Admissions to HIMASP with a Diagnosis of Ischemic Heart Disease. The numbers in parentheses represent the percentage of total adult admissions for that particular diagnosis and year. The numbers outside parentheses represent the actual number of patients admitted.

Dx	Year	1999	2000	2001	2003	2004	2005
410		15(1.57%)	15(1.73%)	11(1.26%)	9(1.29%)	4(0.54%)	18(2.23%)
411		16(1.67%)	9(1.04%)	15(1.72%)	8(1.14%)	16(2.17%)	9(1.12%)
413		4(0.42%)	2(0.23%)	2(0.23%)	0(0.0%)	0(0.0%)	2(0.25%)
Total		35(3.65%)	26(3.00%)	28(3.21%)	17(2.43%)	20(2.71%)	29(3.60%)
Admissions*		958	868	873	699	738	807

^*Exact adult admission numbers for the time periods studied were unavailable. The number of adult admissions during these time periods was estimated based on the average number of adults admitted during the months of October and November between 2005 and 2008.

A total of 46 admissions for the specified coronary events out of 1606 adult admissions were tabulated for the election years 2000 and 2004. This accounts for 2.86% of adult admissions during the specified time frame. The number of admissions for the specified coronary events in non-election years was 109 out of an estimated total 3337 adult admissions, accounting for 3.27% of the admission. Based on the methodology outlined above we estimated that the average percent of patients over the age of 18 years admitted during the time period studied was 81.4% of patients. Thus in election years the proportion of admissions in the time period allotted was 2.86% as compared to 3.27% in the control time frame (see Figure 1). Statistical comparison of admission rates for cardiovascular events on election years vs. non-election years showed no significant difference.

DISCUSSION

The results of this study are noteworthy in that they reveal an unexpected trend towards fewer admissions for acute cardiovascular events during the week preceding and following general elections in Puerto Rico. Although this trend was noted at both of the studied institutions, it was statistically significant only at HURRA. Such a finding would seem to suggest that elections convey some degree of protection against coronary events amongst patients of this institution or that patients delay seeking health care from acute cardiovascular events during this time period. Puerto Ricans are highly involved in the electoral process as evidenced by their high voter turnout (78.04% of inscribed voters in 2008 participated in the general election [9], compared to 61.7% of the voting eligible population in the 2008 US general election [10]. An additional potential explanation for these findings is that patients are more diligent in their self-care during this time period in order to partake in the electoral activities. While the charged emotions surrounding this event would suggest an elevated level of stress, one could also surmise that those who find the proceedings enjoyable would instead be protected against coronary events or their symptoms by the production of endogenous endorphins. These findings, however, would contradict the findings of the previously cited studies conducted around the World Cup, an event which spectators presumably also find enjoyable. A key difference between the World Cup and the electoral process is that the latter extends for a period of time which precedes the date of the election, with events such as campaign rallies and debates occurring throughout the election year. It’s conceivable that the emotional impact of Election Day is diluted by this and that any coronary events induced by stress surrounding the process are distributed throughout the entire year possibly peaking prior to the election rather than near it.

An alternate interpretation, however, is those patients avoid seeking medical attention during the time period surrounding the general election. The desire to partake in the proceedings might cause patients to ignore symptoms that might otherwise prompt a hospital visit. The fact that the difference in admission rates was only significant at HURRA, whose patient population is predominantly composed of uninsured patients or those insured by the public health system, could suggest that socioeconomic factors, such as concern about the potential costs of health care, make patients less inclined to visit physicians around general elections. However, it is worth noting that the HURRA sample size is small and represents a narrow segment of the population. Thus, these results may not be generalizable to other hospitals or populations.

In addition to a small sample size, the study is limited in that it did not analyze mortality data, and thus we do not know whether there is any difference in the magnitude of the coronary event during the time periods studied. Although one may assume that a decrease in admission rates would suggest a decrease in patient mortality, patients who died prior to arrival at the hospital or in the emergency room before a formal admission wound result in an under detection of the event with our methodology.

The study only included patients admitted to the hospital with the specified diagnostic codes used in the study; patients admitted under alternate diagnostic codes and later identified as having experienced an acute coronary event or those evaluated in the emergency room for coronary symptoms but who were discharged from the hospital were not included in the study. Both of these events are potential reasons for a decreased detection rate of this diagnosis. We would, nevertheless expect that a similar effect would be occurring during the control years, offsetting this possible confounding factor. The difficulties in obtaining precise data on adult admissions in one of the institutions along with other demographic information sheds some light into the limitations of doing research using computerized record-keeping software.

In conclusion, there is a significant decline in admission rates for cardiovascular events during the two weeks surrounding the general elections at HU-RRA; a similar trend of less magnitude was seen in the sister hospital at HIMASP. Further studies are needed to assess whether similar decreases occur in other hospitals in Puerto Rico, whether these trends translate into an appreciable difference in patient outcome, and whether any medical or cultural interventions are necessary to modify these trends.