Infection surveillance after a natural disaster: lessons learnt from the Great East Japan Earthquake of 2011

Abstract

Problem

On 11 March 2011, the Great East Japan Earthquake produced a catastrophic tsunami that devastated the city of Rikuzen-Takata and left it without an effective health infrastructure and at increased risk of outbreaks of disease.

Approach

On 2 May 2011, a disease-surveillance team was formed of volunteers who were clinicians or members of Rikuzen-Takata’s municipal government. The team’s main goal was to detect the early signs of disease outbreaks.

Local setting

Seven weeks after the tsunami, 16 support teams were providing primary health care in Rikuzen-Takata but the chain of command between them was poor and 70% of the city’s surviving citizens remained in evacuation centres. The communication tools that were available were generally inadequate.

Relevant changes

The surveillance team collected data from the city’s clinics by using a simple reporting form that could be completed without adding greatly to the workloads of clinicians. The summary findings were reported daily to clinics. The team also collaborated with public health nurses in rebuilding communication networks. Public health nurses alerted evacuation centres to epidemics of communicable disease.

Lessons learnt

Modern health-care systems are highly vulnerable to the loss of advanced technological tools. The initiation – or re-establishment – of disease surveillance following a natural disaster can therefore prove challenging even in a developed country. Surveillance should be promptly initiated after a disaster by (i) developing a surveillance system that is tailored to the local setting, (ii) establishing a support team network, and (iii) integrating the resources that remain – or soon become – locally available.

Introduction

On 11 March 2011, north-eastern Japan experienced one of the most catastrophic natural disasters ever recorded – the huge tsunami that resulted from the Great East Japan Earthquake.¹ In Rikuzen-Takata, in Iwate prefecture, a 16 m-high wave swept across 85% of the city and damaged or destroyed many buildings, including the four-storey city hall, the prefectural hospital, seven of the city’s nine health clinics and all nine of its pharmacies (Appendix A, available at: https://dl.dropboxusercontent.com/u/27176676/onlinedata/BLT13117945/Appendix.pdf). Overall, 1730 citizens – or 7.4% of the city’s population, including two physicians, six public health nurses and many other health-care professionals – were killed or presumed to have been killed by the tsunami.² The city’s health-care system – which had had a shortage of physicians even before the disaster – was left in total disarray. The Iwate prefectural government soon called in health-care support teams from elsewhere in Japan.

By early May 2011 – 7 weeks after the disaster – 16 teams of local or visiting clinicians and one team of public health nurses were providing health care in Rikuzen-Takata.³ At the same time, 15 804 refugees – 70% of the city’s surviving citizens – were still living in the 88 village halls, factories, schools and indoor sports facilities that had been converted into evacuation centres (Appendix A).⁴ Most of these evacuation centres were overcrowded and had inadequate insulation and air conditioning. The city had a working mobile phone network at this stage but the signal was weak in coastal and rural areas. The city’s electricity, water and sewerage services remained disrupted in most areas.⁵

After the tsunami, infection surveillance was urgently required to protect Rikuzen-Takata’s refugees from major outbreaks of communicable disease.⁶ However, it soon became clear that the chain of command among local government administrators and the city’s health-care support teams was too weak to permit effective surveillance.² Approximately 1 month after the disaster, the Iwate prefectural government therefore formed a taskforce to facilitate infection surveillance (Appendix A).^7,8 Between 13 April and 16 August 2011, this taskforce used a specialized tablet computer to collect information on cases of infection from the evacuation centres – about 300 – that then existed in Iwate prefecture. The taskforce, which collected data from a mean of 13.2 centres per day, included a mobile support team. The taskforce’s surveillance data allowed the mobile team to alert refugees to potential disease outbreaks.⁸ However, the data were not made available to most of the other health-care support teams working in the city or even to the city’s Health Service Bureau. On 23 April 2011, nine clinics that had been established – or re-established – after the disaster initiated their own surveillance for influenza, measles, enteritis and scabies in Rikuzen-Takata. This was a response to a recommendation of a health-care team from Kobe – a city that was left devastated by an earthquake in 1995. Although the clinic data were assimilated at daily coordination meetings, they were never systematically analysed.

The main aims of our project were to establish an efficient system of infection surveillance to cover all of the clinics in Rikuzen-Takata – using only the resources that were locally available – and to maintain this system until most of the evacuation centres had been closed.

Strategy and process

A surveillance-specific “working team” of seven volunteers – who had then already been collaborating for about 1 month – was formed on 2 May 2011 (Appendix A). This team comprised a government officer from Rikuzen-Takata’s Health Service Bureau and physicians and paediatricians from Iwate Prefectural Takata Hospital in Rikuzen-Takata or St Mary’s Hospital in Kurume. The members of the team contributed to the project on a part-time basis and had other daily tasks.

Immediately after its creation, the working team spent three days interviewing representatives of the 16 health-care support teams that were then operating in Rikuzen-Takata. Their aim was to assess local conditions and determine the resources that were locally available. Eight of the support teams were based in the Takata or Yonesaki districts of Rikuzen-Takata (Appendix A) and six of the other teams joined twice-daily coordination meetings at Yonesaki. However, most of the support teams provided their services on a stand-alone basis and the chain of command among the teams was poor. All of the support teams encouraged the launch of a unified system of infection surveillance in Rikuzen-Takata. A reporting format for cases of infectious disease that was consistent with pre-existing surveys, minimally burdensome and flexible – in terms of the methods that could be used to pass on the data for assimilation and analysis – was commonly requested. The working team concluded that (i) infection surveillance should be clinic-based, to detect early outbreaks; (ii) the adverse impact of infection surveillance on clinicians’ workloads should be minimized by introducing a simple reporting format, and (iii) the main problems in communication – for example, for the daily submission of completed report forms to the working team and for alerting the relevant health-care professionals about any increased risk of a disease – could probably be overcome by establishing a communication network at the “grassroots” level.

Draft and refinement of protocol

The two infection surveillance systems that had already been explored in Rikuzen-Takata after the tsunami were both based on a reporting format recommended by the United States Centers for Disease Control and Prevention.⁹ The working team drafted a simple reporting format that combined aspects of these earlier systems. To promote the new surveillance project, the working team visited each clinic in Rikuzen-Takata in the company of the existing team of public health nurses – who had all volunteered to support the project. During these visits, time trend data from earlier, preliminary surveillance were presented. Participants – the clinicians working for the support groups in the clinics – were told that they could use any available communication method to report cases of infection – to the Health Service Bureau, a support group coordination meeting or a public health nurse – or to obtain feedback data from the working team. For example, they could use mobile or landline telephones, facsimile machines or the internet, or simply hand a completed paper surveillance form to a public health nurse. The working team closely shared information and strategies with Rikuzen-Takata’s municipal government. The protocol for the project was also reported to the Iwate Public Health Unit to confirm its consistency with the prefectural government’s strategy for disease surveillance.

Launch and operation of surveillance project

A surveillance form was released as a pilot version on 4 May 2011, and as a final version 2 days later (Appendix A) – when the protocol for the project was approved by all 16 support teams in Rikuzen-Takata and comprehensive data collection commenced. Data recorded on the forms by clinicians working in the support teams were collected daily by the Health Service Bureau from four support teams in Takata. Another 10 support teams submitted completed surveillance forms at daily coordination meetings in Yonesaki or via public health nurses. Two other support teams submitted data from the forms daily, using a facsimile machine or a landline telephone.

One member of the working team – using Excel 2007 (Microsoft, Redmond, United States of America) and a tailor-made macro – developed a database format to facilitate the semi-automated analysis of the surveillance data and the preparation of summary findings. Each day, time trend data on the incidence of each recorded infectious disease over the previous two weeks were sent to each support team (Appendix A). These summary results were also displayed at each major base used by any of the support teams. As they visited the support teams, the public health nurses regularly thanked the participants and encouraged further data collection. At the start of the project, the members of the working team each spent up to 50% of their daily work time on the project protocol and its implementation. However, once the project had been running for a few weeks, the daily time that the working team spent on the project fell substantially, partly thanks to the concurrent restoration of landline telephone lines throughout most of Rikuzen-Takata. Subsequently – despite the gradual decline in the visiting workforce – infection surveillance was maintained, at an average response rate of 96.2%, until the end of the project. The project was terminated on 13 July 2011, when most of the evacuation centres had been closed and Iwate Prefectural Takata Hospital had been re-opened, albeit at a temporary site (Appendix A). The quality of the surveillance data that formed the focus of the project was deemed to be generally satisfactory. Records that were incomplete or ambiguous were rarely encountered. The quantity of additional, voluntary information that was assimilated – such as descriptions of specific cases of some diseases – varied significantly according to the support team involved.

Support teams were alerted – by the working team – whenever the working team thought it had detected the early signs of an outbreak of a communicable disease, such as the observation of at least two cases of antigen-positive influenza on each of the previous two days. If the suspected outbreak continued to expand, refugees and other citizens were also alerted, either by the Health Service Bureau – generally via an announcement in the weekly bulletin that was published, in printed format, by Rikuzen-Takata’s municipal government – or by the team of public health nurses. The public health nurses advised the staff working in evacuation centres on how best to isolate patients with highly contagious diseases – such as influenza and norovirus enteritis – within the limited space that was available. For the other communicable diseases that were observed – such as common cold, mumps and hand, foot and mouth disease – the cases, their caregivers and neighbouring refugees were offered preventative measures such as facemasks and alcohol-based hand washes.

Findings from the infection surveillance

During the project, 16 587 patients were seen by clinicians working for the support teams in Rikuzen-Takata. Of these patients, 44.3% were older than 64 years and only 9.6% were younger than 15 years (Appendix A). The most frequently reported problem was respiratory disease (1437 cases), followed by gastrointestinal illness (301), skin or soft-tissue lesions (164) and fever (20). Neuromuscular disorders (7) and jaundice (1) were also formally recorded, whereas cases of some other diseases, such as mumps, were only mentioned in the “comments” section of report forms. The incidences of respiratory illness and gastrointestinal illness were both highest at the commencement of the surveillance. Influenza viral infection was confirmed – with a positive antigen test – in each of 113 patients. Small outbreaks of mumps and of hand, foot and mouth disease were mentioned on the report forms, in May 2011 and June 2011, respectively. None of the recorded disease outbreaks developed into a serious epidemic or pandemic.

Discussion

Health-care systems in developed countries are heavily dependent on modern technology and are, in consequence, very vulnerable to natural disasters such as the Great East Japan Earthquake. The tsunami that struck Rikuzen-Takata in 2011 left no effective system of disease surveillance. It also hampered early attempts to re-establish such a system because it broke the necessary chains of command and devastated the city’s communication network. The failures of these early attempts were, however, soon recognized and – as the city’s whole infrastructure was slowly rebuilt – a new and effective system of disease surveillance was created from the resources that were locally available. In the post-disaster development of a health-care system, the formation of at least one on-site “working team” that can optimize the system for the local setting and facilitate a support team network using the best available resources should be considered.

Post-disaster infection surveillance today

To protect the survivors of natural disasters from diseases,^10,11 the prompt establishment – or re-establishment – of a system of infection surveillance is essential.^12,13 Infection surveillance is known to have played an important role in preventing outbreaks of communicable diseases following the 2004 Indian Ocean earthquake and the 2008 Sichuan earthquake.^13,14 After the Great East Japan Earthquake, primary medical care was rapidly provided by the development of several stand-alone clinics that were run by local and visiting teams of health personnel. The rigorous cooperation between support teams and the effective communication network – that, together, would allow the rapid redevelopment of an effective city-wide system of health care – took longer to develop.

Initiation of infection surveillance

To overcome the lack of an efficient command chain among the health-care support teams that worked in Rikuzen-Takata in the months immediately after the 2011 tsunami, the Iwate prefectural government created a small taskforce and provided it with a tablet computer to facilitate infection surveillance in the evacuation centres.^7,15 This taskforce was able to take advantage of the early creation of a mobile phone network that covered all of the larger evacuation centres. While precious information was collected by the taskforce, the benefit of that information to the health-care support teams working in Rikuzen-Takata at the time was very limited. The taskforce covered too small a population and used an inflexible and relatively ineffective system for disseminating the data to those who would have found them useful.⁸

In an attempt to overcome the taskforce’s limitations, we aimed to establish clinic-based surveillance across the whole of Rikuzen-Takata. As many modern methods of communication remained unavailable at the smaller evacuation centres and clinics, we allowed and encouraged participants to use whatever methods were available to them. Most participants simply passed their completed surveillance forms to a member of the working team or a public health nurse and none used e-mails to submit their data. The clinic-based surveillance system benefited from reliable information provided by medical professionals. The difference between the incidence of antigen-test-positive influenza recorded in the clinic-based surveillance (0.07 cases per 1000 citizens per day) and the incidence of influenza-like illness recorded in the evacuation-centre-based surveillance (0.6 cases per 1000 evacuees per day)⁸ illustrates the degree to which the results obtained with the two approaches can differ.

Resources needed for infection surveillance

Careful assessment of local settings and the development of a support team network are essential for the optimal implementation of post-disaster health-care strategies and, for this, the creation of an active on-site “working team” is recommended. As our working team was not supported by specific funding, its members had to divert a substantial amount of their work time to the project, although they were all formally employed to conduct other activities. However, after the clinic-based system of infection surveillance had been implemented for a week, the members of the working team had to spend much less time on the system. This trend was the combined result of a smooth-running system once any “teething” problems had been resolved and the once-novel procedures had become routine, the semi-automated nature of the system of data-processing, and the concurrent restoration of the landline telephone system in Rikuzen-Takata.

Post-tsunami infectious disease outbreaks

Following tsunamis, an atypical form of pneumonia known as “tsunami lung” has been reported in survivors who nearly drowned.^16,17 We observed an outbreak of respiratory illness in May 2011 but the symptoms were generally confined to a cough.¹⁸ Given that this outbreak did not appear to affect children younger than 5 years – who generally spent more time indoors after the disaster than their adult counterparts – the symptoms might be attributed to dusty tsunami debris in the air.¹⁹ Outbreaks of gastrointestinal illness and influenza were also noted. Precautionary measures and isolation of patients might prevent further cases of these diseases.

Conclusion

Our observations in Rikuzen-Takata indicate that efficient infection surveillance can be swiftly established after a catastrophic natural disaster, even without specific funding or the full-time employment of dedicated staff. Health-care support teams should be allowed to use the best method of communication that remains available. Ideally, an on-site multidisciplinary working team – which can assess region-specific characteristics of the disaster area and establish rigorous cooperation between the various support teams using the best available resources (Box 1) – should be formed.

Box 1. Summary of main lessons learnt.

• Modern health-care systems are highly vulnerable to the loss of advanced technological tools.

• The support teams responsible for re-establishing primary health care following a catastrophic disaster need to cooperate and form an effective but flexible network of communication, data collection and data feedback that is carefully adapted to the local setting.

• The resources that remain – or become – locally available need to be integrated into the surveillance system and to be fully exploited.