Saturday, January 12, 2013

Statistical Shortcuts - No Data

The government has launched a prestigious healthcare scheme for the poor people. It is the Rajiv Gandhi Jeevandayee Arogya Yojana. It aims at giving free treatment to poor people who need any one or more of a list of procedures that have been selected. The civic body took it up for some civic hospitals. The erstwhile Big Boss called us all and told us to work on it.
"We can expect to get 25% of our operative procedures under the scheme" he said. "Is that right?"
"Yes" said a superspecialty Boss, "20%". The Big Boss nodded in satisfaction. I could not see how they arrived at those figures. But then, these guys probably had access to information I did not have.
The scheme rolled on. The civic administration got extremely unhappy with the performance, and Big Boss summoned us. "The performance is very poor. The civic boss wants 10% coverage of all procedures under this scheme today, 20% at the end of the next month and 30% thereafter" he said. Then he started calling out names of specialties and asking their heads for explanation for the dismal performance. When he called for Obstetrics Gynecology and looked at me questioningly and accusingly together, and said "only 0.6% of your cases have been performed under the scheme. What is your explanation?"
We performed more than150 procedures, of which just about 10 were included in the scheme. The scheme was just for 8 districts, not the entire state.
"Sir, I denominator in the calculations is wrong" I said.
"Don't give excuses" Big Boss thundered. The civic boss had probably thundered at Big Boss for this. I shrugged my shoulders and said "if you won't let me say anything, I won't." Big Boss wanted an explanation, so he asked me what it was. I told him what I have written above. "The true estimate is really 0.6% and not 20-25%. When you take out a percentage, you have to consider the denominator as patients whose eligibility is based on the procedure list, not ALL procedures."
"But YOU gave me the percentage as 20%" he thundered. Civic boss must have been really annoyed with the performance and the person responsible for the implementation of the scheme.
"I did not" I said.
"Who did?"
I pointed at the superspecialty boss who had said that. He was speechless. He made some statement that made no sense to me and was not loud enough for other people to hear anyway.
"I am sorry" Big Boss said to me "if I have shouted". I kept quiet. He had used the word 'if' superfluously, and had omitted to say 'wrongly'. Words mattered little to me, intentions did.
"How did the Big Boss and Superspecialty Boss reach the figures 25% and 20% to start with?" someone asked me later.
"That is a statistical technique of generating results without having any data to analyze" I said. "It is often done by people who reach high positions."

Six months later, when a new Big Boss was in position and was reviewing the dismal performance, the same old superspecialty boss said with wisdom "perhaps the estimate of percentage was wrong. Let us collect some data and see what it actually is." The new Big Boss liked the idea. Then this superspecialty boss saw the expression on my face, did some quick thinking, and asked me for any suggestion.

"I had suggested this five months ago" I said dryly. It is to his credit that he kept his expression bland.

प्रशंसा करायचीय, नावे ठेवायचीयेत, काही विचारायचय, किंवा करायला आणखी चांगले काही सुचत नाहीये, तर क्लिक करा.

संपर्क