|
|
In October 2004 the FDA approved implantable RFID chips for the purpose of giving health care providers access to a patient’s medical information. Public debates about patient privacy quickly followed and in 2007 the AMA updated its code of ethics to address the use of medical chips.
When news about possible links to cancer began to surface as well, the stock for VERICHIP CORP (NASDAQ: CHIP), the maker of the implants, began its slide from double digit dollar levels into the realm of penny stocks. It seems little wonder then, when The Health Care Blog recently reported that one of the chip pioneers is ready to sign off.
Yet, Data Management Issues Play a Major Role As Well
The cancer scare and the specter of big brother, nosy neigbors or more nefarious uses probably present the biggest hurdles to adoption. They make for good headlines and everyone can relate to them on some level.
Even if we set these issues aside, the medical chip still faces significant problems. As we wrote back in 2004, setting up a system that makes medical information available via an RFID chip presents significant logistical challenges:
“… in order for the life-saving-promise to work, a couple of things have to be in place:
- a repository that contains all critical medical information relating to the patient
- the information in that repository has to be up-to-date and correct
- someone has to provide information to the repository
- medical personnel needs to have a scanner that can read the chip and access the information
- the “critical” information has to have been collected in the first place
Just imagine what’s necessary to put all these things in place: someone has to create an electronic version of the data, someone has to be in charge of the information to make sure it’s accurate and up to date, there has to be a process for correcting and updating information, all hospitals and emergency responders have to be equipped with scanners, security features and processes have to be added to prevent unauthorized access, and that’s just the tip of the ice berg.”
Data management issues induce yawns in many people, but they are at the core of the bottom line decisions that businesses have to make every day: at some point the investment has to pay for itself. Solving these challenges requires significant funds and resources. The current push to save health care costs via eMR systems may just provide the booster shot that’s needed. With this in mind, it seems likely that the RFID idea is not dead, but will find a different life as part of a smart card.
Additional Reading:
AMA: New ethics policy
Recommendations adopted from reports presented at 2007 Annual House of Delegates Meeting (June A-07):
http://www.ama-assn.org/ama/no-index/about-ama/14449.shtml
Chip Implants Linked to Animal Tumors
By TODD LEWAN
The Associated Press
Saturday, September 8, 2007; 2:04 PM
http://www.washingtonpost.com/wp-dyn/content/article/2007/09/08/AR2007090800997_pf.html
Sometimes counting pills just isn’t enough. When developing forecasts or trying to track product performance one has to consider how a drug is being used and look beyond the mere number of pills or vials on the market.
A good starting point includes looking at standard courses of therapy, average daily dose and similar proxies. A better approach also considers factors such as treatment location, actual patient usage data, insurance coverage, and insights derived through primary research. Each of these attributes takes real patients into consideration and thus can provide a better sense of what is happening in the market place.
These are some instances when patient centric modeling makes sense:
- A mature market with generics and significant off label use
- When products are approved for multiple indications but we are only interested in a sub-set of these indications
- We are trying to assess market potential for a new product or indication without the benefit of a direct competitor
Here is an example of how a patient centric data model may be developed and why a unit centric approach (i.e. “counting pills”) does not always work.
Modeling data around patients involves more work than just counting pills. One has to identify data sources that provide relevant insights and one has to determine where the blind spots in that data are. Since no data is perfect, one also has to figure out how to work around these blind spots.
Not only does such a model have to be developed, it also has to be maintained. The market evolves over time, and today’s assumptions will no longer work at some point in the future. To maintain a model requires regular updates and reviews of the underlying data and assumptions. It also requires frequent “gut checks” against other market intelligence and an understanding of market influences.
Bottom line: patient centric modeling may not always be cost effective or appropriate, but when multiple indications and uses muddy the playing field it really clears things up when we can consider how products are actually being used.
How to reduce the cost of our health care lies at the center of the health care reform debate. It’s no wonder then that “cost containment” became a primary point of focus in the recent ruling by the U.S. District Court in Vermont upholding a Vermont state law which essentially prohibits the use of prescriber data to market and sell pharmaceuticals.
Basically, the VT Attorney General argued that keeping prescriber data from sales reps makes their sales efforts less effective and thus keeps doctors from prescribing newer (i.e. riskier) and more expensive branded drugs – while saving Vermont a significant amount of taxpayer money. The counter argument, of course, states that prescriber data allows sales reps to be more efficient and thus helps reduce the cost of drugs.
The fall-out from major drug failures and headlines questioning the legitimacy of drug prices makes it easy to see why the VT legislature saw it fit to pass the law at issue here. On the other hand, the majority of drugs on the market are not blockbuster material and are not the ones driving current headlines. Smaller companies do benefit from prescriber data that allows them to allocate their more limited resources to areas where their drugs are more likely to be needed.
Increased conflict of interest controls, lackluster product pipelines, the move toward so-called personalized medicine, cost constraints and other industry developments already are changing the way in which drugs are marketed and sold. It will be interesting to see how the legal details continue to play out, but by the time the final rulings are made, the industry may have moved on.
Further Reading
PDF file of the 4/23/2009 ruling by the US District Court for the District of Vermont
http://www.prescriptionproject.org/tools/solutions_resources/files/0040.pdf
The 2009 H1N1 flu scare presents an interesting case study for making critical decisions. Whether to produce a vaccine, how to structure the vaccine and which quantities to manufacture are decisions that pit a potential health disaster against the needs of governments and companies to manage their resources responsibly. From a decision making standpoint each choice represents significant risks and decision makers find themselves in a race against time while searching for the information needed to make good decisions.
In a nutshell, officials and business leaders to have to
- Decide when to act: developing vaccines, producing them and getting them to where they can do the most good takes significant money, time and human resources
- Figure out what to do: doing too little could lead to a major health crises while doing too much could result in wasted resources at a time when our means are already stretched thin
- Find the best time to act: if they act too soon, the wrong vaccine may be developed, significant resources may be wasted and lives may be lost – if action is delayed for too long the proper vaccine may not be ready in time to prevent major loss of live or productivity
- Weigh the relative impact of each decision: regardless of which choices are made, each will have significant effects on people and business – the trick lies in finding choices with the least negative impact.
Regardless of the ultimate outcome, this situation underscores the critical value of accurate, timely and appropriate information. While scientists work around the clock to fully understand this new virus, vaccine manufacturers still have to look at some hard facts – after all, they can only produce vaccines as long as they can stay in business. Some of the critical pieces of information to be gathered include:
- Does this virus present a serious enough threat to warrant the production of a vaccine?
- How many people are really at risk?
- How many doses will actually be needed?
- How long will it take to produce them?
- What are the costs of speeding up production or increasing capacity should this become necessary?
While it may sound inappropriate to worry about bottom lines when people’s lives are potentially at stake, it would be reckless to rush ahead with producing a product that may never be needed.
Further reading:
Officials Face a Tough Decision Over Ordering Vaccine
By JEANNE WHALEN and GAUTAM NAIK
Wall Street Journal (4/30/2009)
http://online.wsj.com/article/SB124105096668271115.html
Swine flu: How serious is the global threat?
James M. Steckelberg, M.D.; The Mayo Clinic
http://www.mayoclinic.com/health/swine-flu/AN02000
As companies grow, so do the demands on their Sales Operations and Marketing Research teams. As new products are launched and sales forces grow, BI managers need to answer ever more questions from ever more people about ever more data and ever more details. The “working smarter” solution to deal with this increasing demand usually involves new skills or new technology or additional resources – and the time and resources needed to implement them.
Working smarter can be accomplished by
- Streamlining processes
- Automating routine tasks
- Reducing errors
- Letting go of less important tasks
- Finding someone with the right skills to do the job.
In this article we will take a look at how to meet increasing demands without simply doing more work.
Streamlining Processes
Most of us easily recognize this scenario: a BI manager is asked to provide a “quick analysis” to answer a burning business question. Sometime later the manager is asked to update the results of this “quick analysis” and before you know it the one-time “quick analysis” turns into a regular report.
Chances are good that the manager initially did not worry too much about how efficiently the analysis was done, as long as the results could be delivered quickly. Chances are equally good, that the process for performing this “quick analysis” can be streamlined to consume less time and resources.
Some ways in which we can streamline processes include:
- Having data in a more useful format
- Pre-calculating certain results and storing them where they can be easily retrieved
- Employing technologies like dashboards and pivot tables to quickly retrieve different views of the data.
If the requested analysis requires significant effort, one should also review the entire process from data collection to processing to producing output. Chances are good that some tasks in that process can be done more efficiently or that by reorganizing the process itself we can save time and effort. If certain tasks have to be repeated over and over again they become a candidate for automation.
Automating Routine Tasks
Our article about comparing a product against competitors with multiple indications provides an excellent case study for automating routine tasks. In that scenario, product units have to be converted into more useful information, taking into account how and where products are used. To account for all this, raw units have to be factored each time new unit data becomes available – a tedious task which is also prone to human error.
Unlike humans, computers excel at processing routine and tedious tasks, so why not give them that job?
Reducing Errors
The more critical the information, the more questions will be asked. This axiom lies at the root of many BI headaches and its impact snowballs when errors occur. Each error we find chips away at the level of trust we have in the information. As BI managers we need to provide the most accurate information we can, given the resources at our disposal.
Automation represents one way in which to reduce errors, but we can do more. Spot checking, quality control steps like tying out sub-totals and checking whether the results “make sense” should be an integral part of every information delivery. Even better, we usually can automate some of these QC steps.
Letting go of less important tasks
The article “Making Time For The Future” addresses this topic in more detail. Suffice it to say here that letting go of less important tasks allows us to spend more time on the more important ones.
Finding someone with the right skills to do the job.
This brings us to the last item on this list: to get all of this done, we sometimes have to look for someone else to do the job – either because we do not have the time or the necessary skills. This raises questions like:
- Can we afford additional headcount or a consultant?
- Is this a long term need or a short term need?
- What exactly should this person do?
- What skills, knowledge and experience do they need to bring to the table?
Answers to these questions depend on many factors, including:
- Budget
- Expected growth of the organization
- To what extent the BI department wishes to combine business skills with IT skills
- Which skills are already available in house
- Whether the organization prefers to do all the work in-house or work with external partners
- Whether the right person can be hired as an employee or a consultant.
The Bottom Line
Successful organizations will require ever more complex information to stay competitive. As the industry changes and sales reps need to blend more and more sales, medical science and account management skills, we have to integrate data from ever more sources. While technologies like dashboards make the presentation of information easier, we are still faced with the tricky task of determining which information needs to be available when and to whom.
This article provides some ideas about how to meet these increasing demands without simply doing more work. It is the second in a three-part series of articles looking at Three Steps To A Better BI Future. In the last article we will take a closer look at how to provide the right information to the right people at the right time.
The three articles in this series are:
Time acts like money: we have to invest a little up front to get much more of it down the road. The initial time investment involves paper, pencil and some brainstorming. We build on this investment with a little bit of research and use the research results to start building a case for stopping things that no longer provide the value they once did. As we stop doing some things we make time for building a better future.
Paper, Pencil and Brainstorming
Let us start the brainstorming session with a simple question: “What keeps us busy every day?” The answer should not be a detailed list of every report the department produces, nor of every research project being conducted. Rather the goal here is to identify where the big chunks of time are going, yet with enough detail so that we can later group our thoughts in meaningful ways. Some meaningful groupings might be:
- by department
- by brand
- by budget line item
- by the type of data being used
- by cost
- by the amount of effort involved.
We may want to start with any or all of these bullets to get the brainstorm going. However, since we are in the process of trying to find TIME for the future, we eventually want to focus on the EFFORT involved.
A Little Bit Of Research
Since perceptions can be deceiving, we need to validate our impressions. At this stage we need to explore whether individual projects really consume as much time as we think they do. This is also a good place to ask whether something still is being used or whether it has the same importance it once did. Some of the more tricky items to review are all the “favors” that we inevitably do for others. At this stage of the game we are trying to:
- verify how much effort is required
- look for low hanging fruit
- search for items that promise big returns.
With some projects it will be obvious that they are important and need to get done, yet we can always explore whether they can be done faster and more efficiently. Options include finding new technology, outsourcing, procuring the co-operation of other departments or hiring someone with more appropriate skills.
For other projects it will be just as obvious that they can be stopped or reduced in size. For everything else, we need the next step.
Building A Case For Stopping Things
Chances are good that eventually we will have picked all the low hanging fruit. Budget constraints, changing business priorities and additional demands may make it necessary to cut back on existing services. And of course, we are still looking to find time for the future!
At this stage we have to weigh the benefits of one project against those of another. We have to weigh the impact each project has on the company overall and we have to parse perceived need from real need. Key questions to ask include:
- How does this make you more effective?
- How does it help you make better business decisions?
- How would it hurt your effectiveness if the information was updated less frequently?
- Without this information, how would you get your job done?
Notice that these questions do not allow for “yes” or “no” answers. They are the starting point for exploring the true value of something that is perceived as a needed service. Only if we and our information customers explore these questions, can we determine whether other options are available or whether the need is as great as we initially thought.
Building a case takes time and effort – and if we reach this stage, we should also weigh the needs of our own department and not just the needs of those we support. To effectively support our information clients we need to also look at how our own service offering needs to evolve to meet changing needs.
Needing Time For The Future
Successful business information managers constantly engage in their own form of R&D, researching and developing solutions that provide timely and accurate information on which to base profitable business decisions. To meet future demands it is vital that BI managers
- Develop expertise with relevant business issues
- Follow market trends
- Learn about new data sources
- Explore new technologies
To be able to do these things, BI managers need to find time in their schedule – or better yet, they need to make time for the future. The tips outlined here should provide a good starting point.
This is the first in a three-part series of articles looking at Three Steps To A Better BI Future. In the next article we will take a look at how to meet increasing demands without simply doing more work.
The three articles in the series are:
Managers who provide information to their sales and marketing organizations walk a tight line between drowning in data details and making sure that the information they provide leads to profitable decisions. The daily fire fights do not leave much time for planning ahead, let alone for envisioning the future. Yet, all successful managers know that the future waits for no one – and that it’s better to be prepared than to catch up. Thus they periodically make time to determine which activities need to:
Of course we all know that this traffic light approach to the future sounds more simple than it is to implement. Implementing “stop-continue-start” often requires multiple doses of diplomacy and the stamina and steadfastness of a referee. Over the course of three articles we will take a look at what it takes to get there.
- Step 1: Making Time For The Future
- Step 2: Looking Ahead – More Data, More Details – Less Work?
- Step 3: Getting There – The Right Blend Of People, Skills And Goals
Pharmaceutical drugs often treat multiple health conditions, and sometimes the dosage varies considerably depending on the patient or the condition being treated. Under these circumstances it becomes necessary to look at how a drug and its competitors are used in real life before one can adequately compare their performance in the market place.
Likewise, comparing products only by unit volume can significantly distort the market picture and lead to decisions based on incorrect information. Let us look at a simple, hypothetical example:
- we have information that 45% of Competitor X’s volume goes toward treating conditions very different from those that our drug treats – how should we account for this in our market analysis?
Some Factors That Influence Market Calculations
Common sense suggests that we should not calculate market shares based on the total volume of Competitor X, since a large portion of that volume represents non-competing business. However, we cannot simply deduct 45% of Competitor X’s volume from the market total. Why?
- The market continually evolves, it may be 45% today, but what will it be tomorrow?
- A significant portion of Competitor X’s volume may flow through channels in which we have decided not to compete – against what volume of Competitor X do we really compete?
- A standard course of therapy using Competitor X may require a different number of units than a standard course of therapy using our drug – how can we tell whether our drug is gaining or loosing market share?
We Need A Better Market Definition
We now realize that we need to make some decisions about how to define our market. In other words, we need to develop some rules for turning total product volume into something more useful. In this simple example, we need to take into account
- how products are used and
- where they are used.
In real life, the factors that need to be considered may be much more complicated.
We also need to gain agreement from business leaders about what these rules should be and what reports will be needed to support their decision making process. Several discussions will probably be needed, because we have to collect and interpret data that will help us come up with suitable rules.
We Need Appropriate Data
This data may come from a variety of sources, both internal and external to the organization. Depending on the complexity of the market and the available budget, we may
- conduct primary research
- buy syndicated data
- acquire data from other sources like distribution partners or the Federal Government.
In this example we may need to look at
- number of units being used to treat various conditions
- product volume in various channels
- which conditions are being treated in which channels.
We Need Customized Calculations
Using our example, we may have determined the following, also hypothetical, rules
- Competitor X requires 4 units per course of treatment while our product requires 3 units
- We have determined how much to discount Competitor X in each channel
- We have chosen to create two market views: a Nation Total and one only for Hospitals
Now that we have determined the rules, we have to set up calculations that apply these rules and turn total product volume into something more useful. But we cannot rest on our laurels – as the market changes, these rules may need to change — this means we periodically have to start this process all over again.
To see how this process is used in real life, take a look at one of our Case Studies.
As companies grow, so does the list of things that need to get done – and at some point things that ought to get done don’t get done because either time or expertise are in short supply. When critical items languish at the bottom of the list for too long, business suffers. Ideally, either the payroll grows or a consultant helps out before things go sour.
Many consultants will be happy to do the work – but picking the right one out the bunch requires some planning and foresight. The journalist’s “What-When-Which-Where-How” approach to story reporting is a good place to start:
- What needs to get done?
- Is critical information lacking?
- Is there too much data, but too little useful information?
- Does conflicting data have to be reconciled?
- Does the business need a new way of looking at things?
- When does it need to get done?
- One time only
- Periodically (daily, weekly, monthly, etc.)
- Which skills or types of experience are required?
- Business analytics
- Industry experience
- Data skills
- IT skills
- Beginner vs. Expert skills
- Where should the consultant be located?
- Are face to face meetings required?
- Can the work be done remotely?
- Is access to company data/IT resources required?
- How should the work be done?
- How much team interaction is needed?
- Who should manage the project?
- Can work results be transmitted electronically?
- What types of presentations and reports are required?
Each situation has its own idiosyncrasies, of course, and this list can be modified and expanded as needed.
Relationships Rule!
Earlier we said that this list is only a starting point because the secret ingredient to success is the relationship between consultant and client. If client and consultant cannot get along, the project sits on shaky ground – no matter how proficient, competent and talented everyone is.
The bottom line:
the Right Expertise + the Right Relationship = Success.
You are working late on a Friday, you are facing another week on the road to conduct physician interviews and you have a choice:
- sacrifice the weekend to the marketing gods
- or tell the brand team that their report will have to wait a few more days – maybe a week.
Since neither option promises a happy ending, some wishful thinking winds its way to the top of your mind: what if someone else could dig into the data, come up with squeaky clean summaries – and, presto, provide you with a weekend unencumbered by crazy deadlines?
Let’s face it: we cannot always do something about the crazy deadlines, but we can find competent help – someone with experience, someone who understands what needs to get done, someone who can do the job just as well, or better, than we can.
If you are looking for a person to help with your marketing research work, you might start with deciding what this person should do.
- Maybe they can handle some, or all, of your secondary data reports – so you can focus on primary research
- Maybe they can compile data from various vendors – so you can get a more complete picture of the market place
- Maybe they can come up with custom metrics in a market that is difficult to measure
Only you know what help you and your organization need, whether you can afford the extra head count or whether a consultant would be a better choice. You may want to consider hiring a consultant if:
- the project requires expertise that is unavailable in-house
- you do not have enough on-going work to justify a full time employee
- you need help with a short-term project
Once you have decided to hire a consultant, you need to develop selection criteria – which is a topic large enough to deserve its own blog post: How To Find The Right Consultant.
|
|
