People are harmed in the hospital when they fall trying to get out of bed and sustain a head injury or broken hip, pick up new infections that may be very difficult to recover from, are given the wrong drug and it creates a serious problem, and so forth.

In May, the federal government published great news comparing harm done by the health care system to Medicare patients in hospitals in 2010 to harm caused in 2012. The report said that harm decreased by 9 percent over those two years. Roughly 15,000 fewer people died as a direct result of care they got in the hospital, and about 560,000 fewer people experienced harm that injured but didn't kill them.

In addition, about 150,000 fewer Medicare patients (mostly age 65-plus) were readmitted to the hospital within 30 days, an 8 percent reduction. When people are readmitted so quickly, often it means that preventable problems were not avoided.

That's excellent news, isn't it?

Yes, it is. Saving all those people from harm or death - or even simply from a return hospital stay - is cause for celebration.

But what else do those numbers tell us? If a 9 percent reduction in harm over two years translates into 15,000 fewer deaths, then 91 percent of the original number still died as a direct result of harm caused by the care they received - about 152,000 people in two years. If 560,000 fewer people were harmed but not killed by the care they got, then nearly 5.7 million people - the other 91 percent - were still harmed. And about 1.7 million Medicare patients were still readmitted to the hospital within 30 days after they were discharged.

Keep in mind that these numbers include mostly people age 65-plus, because they typically reflect only cases involving people enrolled in Medicare. That is, health care also injures or kills a large number of people under the age of 65, and this harm is not counted in the above numbers. Different studies count and measure differently, so the conclusions about total numbers harmed or killed may be different.

But regardless of what groups of patients are studied, and what the exact totals are, the damage caused is staggering. Researchers typically conclude that roughly 40 to 70 percent of the harm and deaths directly caused by health care could be prevented. Similarly, about half of hospital readmissions are considered preventable.

So while the progress made in the last couple years is very welcome, much work still needs to be done. The very regrettable fact is that so much more could be done so much faster; the critical first step is a decision on the part of people running hospitals that it's important to reduce this harm - that your life and well-being truly matter.

Consider this: for five years, 2007-2011, the percentage of Medicare patients readmitted to the hospital stayed the same. No improvement at all was noted. Then suddenly in 2012 and 2013, the rate dropped by 8 percent. That's a big improvement. What happened? Did everyone suddenly get smarter about making health care better? No, some people would say that the answer is much simpler: the government started cutting payments to hospitals that had too many patients readmitted. With income at risk, hospitals started to move away from, "It's not my problem," to "We better do something or we won't get paid as much."

The government, working with most of the hospitals in the country and many, many professionals in health care, created the Partnership for Patients a few years ago to help drive improvements in care - in fact, one of their goals (with a deadline of the end of 2014) is to reduce by 40 percent preventable complications of hospital care. Another is to reduce the percentage of people readmitted to the hospital within 30 days by 20 percent, again by the end of 2014.

While those goals are aggressive and may not be met, they do serve to show that many, many experts believe that huge improvements can take place quite quickly. While it's not as simple as waving a magic wand, methods for making the needed improvements are well known. Driving the required changes requires focus and persistence, but it can be done.