|
|
1 | (14) |
|
Digital healthcare is much riskier than we think, but it can be made far more effective and much safer. |
|
|
|
This book splits up the action into stories of problems, the solutions, and then the better future we can reach. |
|
|
Part I Diagnosis o Riskier than you think |
|
|
2 We don't know what we don't know |
|
|
15 | (10) |
|
For thousands of years, healthcare was held back because we couldn't see and didn't understand the germs making us ill. |
|
|
|
Today, healthcare is being held back because we don't see computer bugs, and we don't understand the risks caused by them. |
|
|
|
|
25 | (8) |
|
Cat Thinking explains our love of all things digital. |
|
|
|
Our hormone-driven love of technology overrides objective thinking. |
|
|
|
Thinking that computers are wonderful, we feel we don't need to worry about looking for rigorous evidence that they are safe and effective. |
|
|
|
|
33 | (16) |
|
Look carefully for them, and you'll uncover lots of stories of digital healthcare bugs. |
|
|
|
This chapter has lots of examples of buggy digital health. |
|
|
|
|
49 | (12) |
|
Denise Melanson died after a calculation error that led to a drug overdose. |
|
|
|
What can we learn from the incident? |
|
|
|
|
61 | (8) |
|
Swiss Cheese famously has holes, which can represent the holes and oversights that lead to harm. |
|
|
|
The Swiss Cheese Model has become a powerful way to help think more clearly about errors and harm. |
|
|
|
7 Victims and second victims |
|
|
69 | (12) |
|
When patients are harmed, staff often get blamed -especially when nobody realizes how digital systems can go wrong and create the problems. |
|
|
|
8 Side effects and scandals |
|
|
81 | (28) |
|
We accept that medical interventions like drugs and X-rays have side effects. |
|
|
|
It makes a lot of sense to think of digital healthcare as having side effects too, and therefore it should be evaluated and regulated as carefully. |
|
|
|
9 The scale of the problem |
|
|
109 | (12) |
|
We don't know how many people are dying or being harmed from errors in healthcare, let alone those caused by digital errors. |
|
|
|
What are the facts, and what can we do about it? |
|
|
|
10 Medical apps and bug blocking |
|
|
121 | (16) |
|
Medical apps are very popular, but they are as prone to bugs as any other digital system. |
|
|
|
This chapter gives some typical examples and begins to suggest solutions. |
|
|
|
Like all digital healthcare, apps could be designed to block bugs and avoid the harms that follow. |
|
|
Part II Treatment: Finding solutions |
|
|
|
137 | (8) |
|
The car industry has made cars much safer since the 1960s. |
|
|
|
What can we learn from car safety and from why car safety improved to help improve the safety of digital healthcare? |
|
|
|
|
145 | (6) |
|
Focusing on the bad stuff is the traditional Safety One approach. |
|
|
|
Safety One is unconstructive. |
|
|
|
Instead, Safety Two means focusing on doing more good. |
|
|
|
Safety Two emphasizes doing more good things and therefore squeezes out the bad things. |
|
|
|
13 Computational Thinking |
|
|
151 | (26) |
|
There's a lot more to digital health than being excited about digital computing. |
|
|
|
We need to learn how to think computationally to take full advantage of digital. |
|
|
|
Computational Thinking is the mature way to think about computing - and digital healthcare. |
|
|
|
(Don't forget that the computer chip ere' means that this is a more technical chapter.) |
|
|
|
|
177 | (16) |
|
Drug doses and other forms of patient treatment require detailed calculations. |
|
|
|
Calculation errors are one of the most common types of error and they could be reduced in many ways. |
|
|
|
Calculators themselves ignore errors, and they should be fixed if they are going to be used in healthcare. |
|
|
|
|
193 | (8) |
|
Software warranties generally deny all liability for problems. |
|
|
|
Manufacturers and developers should be required to be more accountable. |
|
|
|
Everyone needs to be constantly curious about improving systems and reporting problems. |
|
|
|
16 Regulation needs fixing |
|
|
201 | (10) |
|
Digital healthcare needs much better regulation - and regulation needs to keep up with the unique issues of digital healthcare. |
|
|
|
Better regulation is a Safety Two approach: regulate for better processes to stop things going wrong. |
|
|
|
|
211 | (14) |
|
Cybersecurity is a serious problem for all computers and digital systems. |
|
|
|
In healthcare, patient safety is paramount, but in the wider world, security has a higher profile than safety. |
|
|
|
Both have problems caused by poor programming and all the design and development processes that precede actual coding. |
|
|
|
|
225 | (20) |
|
Who is profiting from our data? Is Artificial Intelligence (Al) the solution to better healthcare? |
|
|
|
|
245 | (14) |
|
Interoperability - or, rather, lack of interoperability - is a besetting problem in healthcare. |
|
|
|
We need digital to work seamlessly - to interoperate - across all specialties, disciplines, and healthcare institutions (taking due account of privacy, cybersecurity, and so on). |
|
|
|
It requires new thinking to get there. |
|
|
|
|
259 | (18) |
|
Understanding how humans make mistakes in predictable ways is the first step towards making fewer mistakes. |
|
|
|
This applies to clinicians, and most especially to programmers whose mistakes end up as bugs affecting thousands of users. |
|
|
|
|
277 | (24) |
|
Understanding how computers can avoid bugs and mistakes is the first step toward programming safer and more dependable systems. |
|
|
|
This chapter introduces some important software engineering ideas that can help make safer digital systems. |
|
|
|
|
301 | (12) |
|
However good a computer system is, it still needs to do what's needed - not what we think is needed. |
|
|
|
User Centered Design finds out how people really use systems, and how to improve their experience and reliability. |
|
|
|
|
313 | (12) |
|
User Centered Design means finding out how systems are used with their real users doing real tasks. |
|
|
|
The insights from working with users leads to design insights and ways to improve the systems. |
|
|
|
These ideas are formalized in the important idea of iterative design. |
|
|
|
|
325 | (12) |
|
Developers and programmers need no qualifications to develop digital healthcare systems. |
|
|
|
We need to develop a qualification structure for digital healthcare, and do much more research on digital safety. |
|
|
|
Both will have a huge impact on frontline safety. |
|
|
|
|
337 | (10) |
|
Why is poor-quality software so widespread? Simple bugs might seem trivial, but they are very common and don't help patient safety - they make everyone inefficient and error-prone, if nothing else. |
|
|
|
Health would improve if we paid attention to digital details. |
|
|
|
|
347 | (20) |
|
Aviation safety relies on getting very complex engineering right, and it's getting safer and safer. |
|
|
|
What can digital healthcare learn from aviation and aviation engineering? |
|
|
|
27 Stories for developers |
|
|
367 | (16) |
|
We should program better so that digital healthcare gets safer, which is a Safety Two approach. |
|
|
|
Formal Methods is widely used in safety-critical industries, but not often enough in healthcare. |
|
|
|
Here's why Formal Methods is needed, and how it works. |
|
|
|
|
383 | (18) |
|
Although it helps everything else, Formal Methods isn't enough on its own. |
|
|
|
Thorough testing is essential to ensure things really work well, especially when things are going to be used in complex environments like healthcare. |
|
|
|
|
401 | (16) |
|
Let's have a reliable way of clearly seeing how safe systems are, so that we can make choices based on evidence and improve safety. |
|
|
Part III Prognosis: A better future |
|
|
|
417 | (20) |
|
We've emphasized problems and solutions to problems, but of course digital can do some fantastic things too. |
|
|
|
This chapter collects some positive stories about digital successes and how digital can transform lives. |
|
|
|
|
437 | (18) |
|
The horrific COVID-19 pandemic has forced healthcare systems to innovate in digital health. |
|
|
|
Some changes have been amazing, liberating patients, and protecting healthcare staff - but some have been rather worrying. |
|
|
|
|
|
32 Living happily ever after |
|
|
455 | (16) |
|
There's a future world where digital healthcare works, and works well. |
|
|
|
|
|
|
471 | (26) |
|
This book isn't the end of the story about digital healthcare, and its problems and solutions. |
|
|
|
This chapter on recommended reading gives lots of suggestions to help take your thinking further. |
|
|
|
|
497 | (56) |
|
Supporting this book are over 500 notes and references on digital healthcare and patient safety incidents. |
|
|
|
These notes cover media stories, peer-reviewed cutting-edge research, as well as national and international reports. |
|
|
|
Notes in bold are especially good sources for further reading. |
|
|
|
35 Healthcare openness and acknowledgments |
|
|
553 | (4) |
|
Traditional healthcare views of confidentiality are challenged by digital healthcare. |
|
|
|
This chapter also includes heartfelt thanks to all the patients and healthcare staff and others who've told their stories and brought this book to life. |
|
|
Harold Thimbleby |
|
557 | (2) |
Fonts for cancer |
|
559 | (2) |
Index |
|
561 | |