The best businesses know what their customers want and work hard to supply it to them. Toyota just this past sales quarter became the number one car sales company in the world. In typical fashion, they issued a statement saying they didn’t care if they were number one so long as their customers were happy.
In healthcare, does it make a difference if we have happy patients and employees? Yes it does. Some providers are in very competitive markets, such as lasik eye surgery. Good service and care from these providers means good word of mouth advertising and more customers. Perhaps you are not in such a competitive field and really see no reason to work at pleasing your clients and patients. I beg to differ, though. I think you would agree with me that a patient who trusts you will be more likely to comply with instructions. One way to gain such trust from a patient is to work hard to provide the best service that you can and, thus, please the patient.
Merely thinking that you provide good service is not enough, however. You should measure patient and client satisfaction. One way to do this is with a survey. That may seem like a simple thing to do, some of you may think, but there are many pitfalls in doing surveys. One such common pitfall is the failure to insure randomness in the collecting of data. Surveys in which data is not randomly collected are basically useless, if not misleading. Using the results from such surveys can lead to poorly invested time and money in products and services which are useless to your practice if not damaging. Fortunately, there are several good techniques that are easy to use to collect good data.
If you are a primary care provider, I suggest that you ask every nth patient to complete a survey. For instance, you might hand a survey to every tenth patient a day, every other business day for a month. This is called systematic sampling and is commonly used in manufacturing. It would be best to have the client to fill out the survey right after receiving service so as to evaluate that particular visit. This technique works well for walk-in clinics too. I have read in one recent magazine for primary care providers an article suggesting sending surveys to a randomly selected group of patients with postage return envelopes provided. Unless the percentage of returned surveys is high, say 60% or more, I would suspect the results. In fact, in most such surveys the return rate is usually around 10-15%. This leads to skewed results as usually only patients with strong feelings or opinions return such surveys.
If you are involved in prevention activities for a hospital then finding whether you are successful is very difficult. How do you measure something that doesn’t occur, such as Type II diabetes in a population that you are serving? The best way is to take a random survey in the neighborhoods that you serve. That is, send representatives door to door with surveys and have residents fill them out, much like a census taker. A good way to insure randomness is to divide the area you want to survey into approximately equal geographic areas (assuming that population density is fairly constant) and randomly select households in each area so that the same number are surveyed in each area. This may seem difficult but sample sizes need not be too large to get good results. In fact, getting 300 surveys randomly chosen in a population will yield very good results. Getting a 1000 or more yields excellent results. If you want to examine results by some demographic value, such as age or gender, then you should increase your sample size. For instance, if you want to get good results based upon gender, you should get 600 instead of 300 and get equal sample sizes for male and female. Such sample sizes will yield results that are accurate within 3% for basic “yes or no� questions. In a grant application that I am helping prepare for prevention activities in Kent County, Michigan, we plan to use a sample size of 1200-1500 for a survey that we intend to do. This technique also works well for providers in competitive services to see if their marketing strategies are effective. This is an example of a stratified sampling.
Some providers may be able to use simple random samples. This is similar to drawing names from a bowl in which the slips with the names are well mixed. You could number all your clients or patients and randomly select numbers with a calculator or computer and generate a list to survey. This strategy works well for insurance companies wanting to survey physician groups to see if their HEDIS requirements are met. Savings could be generated in this approach if the collecting of information is well organized, such as planning driving routes to the groups that minimizes driving time and distance. This survey technique also works well for phone surveys of existing patients in a primary care setting.
Unlike the consistent application of evidence-based care in healing, the collection of good data to measure satisfaction of clients and patients must involve randomness. One of the three techniques illustrated above�"simple random sample, stratified sample, or systematic sample�"should meet the needs of most healthcare providers and insure accurate results. With the results in hand, healthcare providers should be able to improve the delivery of services, thus increasing market share and patient health.
Tidak ada komentar:
Posting Komentar