Portal de Periódicos da CAPES

Long-Term Changes in Stroke-Related Hospital Admissions After the Fukushima Triple Disaster

2015; Wiley; Volume: 63; Issue: 11 Linguagem: Inglês

10.1111/jgs.13812

1532-5415

Stuart Gilmour, Amina Sugimoto, Shuhei Nomura, Tomoyoshi Oikawa,

Disaster Response and Management

To the Editor: The town of Minamisoma in Fukushima, Japan, lost approximately 1% of its population to the Great East Japan tsunami on March 11, 2011,1 and was also affected by the Fukushima nuclear plant accident. After this accident, approximately 80% of the city population was evacuated,1 and citizens who subsequently returned were older, changing the distribution of disease risks in the town.2 Although the short-term effect of the Great East Japan tsunami on cardiovascular mortality is understood,3, 4 longer-term effects on hospital admissions and mortality are not yet understood. The current study explored the long-term effect of this disaster on stroke-related hospital admissions. All inpatient medical records for the neurosurgical unit at Minamisoma Municipal General Hospital between September 11, 2007, and December 31, 2013, were audited; individuals whom the admitting doctor diagnosed with cerebral infarction, subarachnoid hemorrhage, or intracranial hemorrhage were classified as having had a stroke. Monthly resident population records from the Minamisoma municipal government, adjusted in the 6-month period after March 11, 2011, using data from a separate study that collected an evacuation history for 8,281 subjects attending an internal radiocontamination screening, were used.5 The data were divided into pseudo-months that started on the 11th of each month and extended to the 10th of the following month, setting the population for each month to be the population registered with the Minamisoma municipal office at or near the start of the given pseudo-month. For example, the population for the pseudo-month starting on May 11, 2012, was set to be closest to the registered population value on or after May 11, 2012, and before June 10, 2012. Crude monthly admission rates and admission rates directly standardized against the 2010 population were calculated for each pseudo-month. Data were analyzed using Poisson regression restricted to those aged 35 and older because most strokes occur after this age and adjusted for age group, sex, type of stroke, time, an indicator variable for the earthquake, and their interactions. All analyses were adjusted for population and conducted using Stata/MP 13 (Stata Corp., Chicago, IL), with backward stepwise removal of nonsignificant interactions. Ethical approval for this study was obtained from the Tokyo University institutional review board (codes 24–40–1011 and 23–46–0113). There were 918 stroke-related hospital admissions during the study period, 460 of which occurred after the earthquake (13.9 admissions per month; crude rate 15.2 per 100,000 population) and 458 before the earthquake (10.7 admissions per month; crude rate 33.1 per 100,000 population). This increase in rates corresponds to an age-, sex-, and time-adjusted greater relative risk of stroke admission of 1.62 (95% confidence interval = 1.23–2.14) after the earthquake than before. Figure 1 shows the monthly time trend in crude and directly standardized admission rates before and after the earthquake. This is the first study to assess changes in stroke-related hospital admission rates after a major disaster over the short and the medium term, adjusted for population changes due to the evacuation. It is unique in estimating changes in the population from an unrelated radiation contamination testing study and adjusting rates accordingly, making it the first study to assess rates of cardiovascular events after a disaster that properly adjusts for evacuation. Although previous studies found short-term increases in incidence of cardiovascular events,6 the current study found that stroke-related hospital admission rates increased and remained high for 2.5 years. These changes are unlikely to be due to health service reorientation, because the same hospital and neurosurgical staff served Minamisoma continuously for the entire period of the study. This increased risk of stroke-related hospital admissions has occurred at a time when Minamisoma faces major changes in its health workforce because of postearthquake relocation. The number of doctors in the town has fallen by approximately 10% and the number of nurses by 25%7 at a time when absolute numbers of strokes are unchanged. Many young families have moved out of the area, and long-standing patterns of intergenerational caring have changed,7, 8 further complicating quality of care for elderly people who are most at risk of stroke. Major changes are needed in the orientation of health services and management of risk factors for stroke if the town is to manage this increased rate of strokes properly. This study has implications for disaster response and recovery beyond the town of Minamisoma. Planning for the immediate aftermath of major disasters also needs to take into account the likelihood of increased acute sequelae of chronic disease, and long-term recovery plans for disaster-affected areas should include careful attention to interventions to improve management of noncommunicable diseases and risk factors for stroke to properly care for elderly populations in the aftermath of major disasters. We would like to acknowledge the assistance of graduate students Saki Narita, Maaya Kita, and Yi Liao for their support in gathering and auditing medical records. Conflict of Interest: None. This study was funded by a Toyota Foundation Great East Japan Earthquake Special Policy Development grant. Author Contributions: Gilmour, Nomura, Oikawa: study concept. Gilmour, Oikawa, Nomura, Sugimoto: data collection and analysis. Gilmour: writing first draft. All authors: content editing. Sponsor's Role: The Toyota Foundation had no role in the design, analysis, or writing of this study and no influence on the decision to publish.