Reducing healthcare costs with wireless technology


Medical authorities are crying out for practical wireless health monitoring devices that patients can wear while living normal lives at home. By the innovative application of body worn Wireless Sensor Networks (WSNs) – using inexpensive, ultra-low power (ULP), interoperable, interference immune wireless sensors connecting to a real time display like a watch, then a home computer and from there to the Internet – patients could avoid or reduce hospital stays while still being in constant contact with their healthcare providers.

The wishes of healthcare systems are hardly surprising considering that the August 2008 ON World Inc. report (“Wireless Sensor Networks for Healthcare”) concludes the use of body worn WSNs could save the health care industry US$25 billion by 2012. Moreover, the savings are likely to become even greater if predictions of a rise in chronic diseases become reality.  

But even with this glaring economic advantage, wireless technology is not being used in health monitoring applications today. The question is – why not?

Moving healthcare into the home

It’s not just the budgets of healthcare providers that would benefit from moving patients out of hospital. The patients themselves benefit from convalescing in their own homes – safe in the knowledge that if there are any problems, the doctor is immediately notified. Newly sick people benefit because there are more likely to found an empty hospital bed. Furthermore, community nursing services can be reduced because there is no requirement for nurses to drive around in order to take routine measurements such as blood pressure, pulse or glucose levels.

And then there are the potential benefits to a “greying” population. Traditionally, elderly people are moved out of their homes and into sheltered accommodation when they become infirm. Unfortunately, this decision is often taken in the interest of the authorities rather then the individual because it’s far easier – although more expensive – to monitor the health of elderly people collectively in an institution rather than in their own homes. This can mean a relatively healthy and capable elderly person – who doesn’t want to move – is “encouraged” to leave home.

Wireless monitoring would allow the healthcare authorities to maintain a high level of vigilance over the elderly while allowing them to remain in their homes as long as possible. This is a recipe for a dramatic reduction in costs and a happier patient. The wireless sensors could be used to monitor vital signs, but also indicate if, for example, the elderly person has fallen over.   

Bluetooth meets the challenge

Wireless technology is far advanced and has pervaded almost all sectors with the notable exception of medical. This is perhaps not surprising given the stringent demands of the medical community. These demands are driven by the need to guarantee that a chosen technology does not increase the risk to a user’s health and is totally reliable.

So what would be needed by a wireless technology to meet the challenges of medical monitoring? Here’s a summary:

  • Interoperability  – an open standard is vital so that products from different manufacturers can communicate with each other;
  • Sensors – these need to be accurate and reliable, incorporate simple pairing, be plug & play and feature auto-recovery;
  • Ultra low power (ULP) with long battery life – sensors require a low power RF radio with streamlined protocol so they can run for months or even years on a coin cell battery;
  • System and device security – transmission of data must be safe and secure to keep medical data confidential;
  • Distribution network – sensors need to communicate with services such as the Internet and the cellular network so that information can be relayed to remote health practitioners;
  • A compelling case for adoption – healthcare institutions are very conservative and need a convincing argument to take up new technology.

Bluetooth low energy meets all of these requirements. For example, the protocol stack is small so the radios consume ultra-low currents when transmitting or receiving (and can hibernate in “sleep” states requiring just nanoamps) and Bluetooth low energy supports AES encrypted wireless communication.

Moreover, Bluetooth low energy will be an open standard ensuring that sensors from different manufacturers establish communication quickly and easily. And because Bluetooth low energy builds on the legacy of Bluetooth wireless technology, it will be easily able to form Personal Area Networks (PANs) comprising several sensors – for example measuring arrhythmias, blood pressure and oxygen levels – communicating with a single “master” device.

What is Bluetooth low energy?

The Bluetooth low energy specification details a short-range RF communication technology featuring ultra-low power consumption, a lightweight protocol stack and integration with Bluetooth wireless technology. (However, it is important to note that Bluetooth low energy will not communicate with legacy Bluetooth chips – fitted to devices such as mobile phones and PCs – adhering to the current v2.1 + EDR standard or older versions. Communication will require Bluetooth chips to be revised to include additional circuitry and software to ensure compatibility with Bluetooth low energy. It is expected that this revision will become common in “dual-mode” devices as the addition of Bluetooth low energy to existing Bluetooth devices requires minimal effort.) The full specification is currently being drafted for release in the second half of 2009.

Bluetooth low energy will be power and cost optimised to provide compact, low-cost and ULP transceivers for PUID (personal user interface devices such as watches) sensors for sport, wellbeing and health, remote control, proximity, mobile phone accessories and HID product categories. What’s more, Bluetooth low energy will be available as an open standard, encouraging many vendors to manufacture the chips, thus ensuring the multiple sources of supply health authorities require.

According to the Bluetooth SIG’s provisional figures for single-mode Bluetooth low energy silicon, it has two performance characteristics that meet the ULP requirements for health monitoring wireless sensors: a modest peak current requirement and wide bandwidth. The Bluetooth SIG says the chip’s peak current consumption will be less than 15mA when transmitting at 0dBm (sufficient for a range of up to 10 metres while obeying the authorities’ power restrictions for operation in the licence-free 2.4GHz band) and slightly less than this when receiving. (See Table 1.)

Table 1: Provisional specifications of Bluetooth low energy chips  

In addition, the Bluetooth SIG says Bluetooth low energy chips will have a bandwidth of 1Mbps. This bandwidth has been carefully chosen because years of field experience with ULP proprietary technology has shown that 1Mbps is the optimal trade-off in exactly the kind of wireless applications Bluetooth low energy wireless technology will target. The trade-off is between transmit power – which goes up with increasing bandwidth – and duty cycle – which decreases with increasing bandwidth for a given amount of data. For example, Bluetooth low energy chips running at 1Mbps only have to transmit at the peak current of 15mA for one quarter of the time of a typical IEEE 802.15.4 radio running at 250kbps and a transmission current of 28mA, to send the same amount of data. (Although some lower power, higher bandwidth IEEE 802.15.4 radios are available.)

Putting the RF transceiver into a deep sleep mode if it is inactive for long periods can further reduce power consumption in the wireless sensor. While figures for Bluetooth low energy’s sleep modes have yet to be released, a comparison with a similar ULP RF proprietary technology shows it’s likely a ‘stand-by’ mode will consume tens of microamps and a deep sleep (or ‘power down’ mode) will consume perhaps 900 nanoamps. This level of power consumption is low enough for a pair of AA batteries to last for months or years. (To put this in perspective, the self-discharge current of an AA battery is around one microamp.)

Figure A: Bluetooth low energy wireless technology features dual-mode and single-mode implementations

A Bluetooth low energy chip will be able to wake up, for example, every 10ms, listen for transmissions (consuming around 15mA for very short period) and then return to a deep sleep state while maintaining an average current consumption as little as some tens of microamps.

Data to the doctor

Bluetooth low energy’s winning advantage over rival wireless technologies is in how the data generated by the sensors will get to the doctor. This is because the sensors will be able to communicate directly with the ubiquitous Bluetooth chip in a mobile phone (providing the phone is using one of the enhanced Bluetooth chips adhering to the low energy specification as noted in the section above). The mobile phone will be capable of acting as the master in an ad hoc PAN network of sensors around the body, and ensuring secure communications. Best of all, virtually everybody owns a mobile phone.

The computing power of a mobile phone means that sensors can send “raw” data – because the mobile phone can do any computation required – simplifying their design and saving power. The mobile phone will be able to store the data and monitor whether vital signs are within prescribed limits. If something looks wrong, the mobile phone can automatically send a warning via SMS to the doctor’s mobile phone.

Not ready for healthcare yet

While Bluetooth low energy is the only wireless technology capable of meeting the demands of the medical community for reliability, security and interoperability, there are challenges to overcome. For example, while it’s an impressive achievement, the world’s cellular network is not perfect. Coverage is less than total, especially in countries with large tracts of sparsely inhabited land like Australia, Canada or Norway (and it’s unlikely anyone will provide billions of dollars for a 100 percent reliable dedicated “healthcare cellular network”).

And even where the network is fully developed calls can be dropped. That’s annoying when talking to a friend, but it’s a bit more serious for a heart disease sufferer whose wireless sensor is trying to contact a doctor because it detects an irregular cardiac rhythm. In the home, call integrity could be assured by relaying the sensor’s data via a wired phone, but that’s not a solution for a patient that demands full mobility.   

Furthermore, the weakest link in the communication chain is actually the 2.4GHz link between the sensor and the mobile phone or landline. This is not because the electronics are unreliable, more because the radio signal is subject to the laws of physics and hence can be attenuated by obstructions; in the worse case a person sitting or lying on the sensor will break the link because radio signals of this frequency can’t pass through the body.

Apart from the technical challenges for wireless connectivity in a heath monitoring application, the medical community needs time – probably years – to thoroughly test it in what is, after all, a potentially life-or-death application. And there are a few ethical questions to be addressed too; for example, before the doctor sends out an ambulance he will want to be positive that 73 year-old Martha’s blood pressure and pulse rate have risen rapidly because she’s ill, not because John, the handsome widower from next door, has come for tea.

All this means you won’t see patients going home equipped with an array of Bluetooth low energy sensors just yet – but one day in the not-too-distant future it’ll be routine.

About the author: Alf Helge Omre is Business Development Manager, Bluetooth low energy wireless technology with Nordic Semiconductor. Nordic Semiconductor is a leading member of the group developing the Bluetooth low energy wireless specification. The company expects to be among the first to market standalone devices meeting the specification. For more, go to