Final Blog Installment

For the past week we did the following tasks:

First, we presented in front of the class on Tuesday.

Second, we went through the feedback that we received during the presentation as a team.

Third, we revised our solution as a team.

Fourth, we updated our high level diagram.

Fifth, we updated our low level "detailed" diagram.

Sixth, we updated our high fidelity wireframes.

Key Screens and Presentation (11/29 - 12/1)

For the weekends, we mainly worked on refining our key screens and our final presentation. We figured out our demo scenario and adjusted our screens to that specific scenario.

User Experience Flow and Screens (11/26)

First we started simplifying our process flow to make a user experience flow to present user interaction with our tool.

Then we started sketching our demo screens for our final presentation and report. We decided to choose the situation where a new user create an account and conduct a doctor search. Our key screens for the demo include the user profile, the search screen, the result screen, and the doctor profile.

Update Work Flow (11/24)

We started our meeting by list all the to-do items for the rest of this project in the Google doc.

Then, we worked on modifying our process flow to incorporate feedback from last Thursday.

 

 Our updated process flow looks like the following:

 

Also, we worked on our search result screen. 

Flow Model and Others (11/23)

After we met with Linda, Karen and Irish (from Premera) yesterday, we started today by redefining our solution system. We started with Chris presenting his thoughts on how the system should work. We also considered an approach similar to dating website, but then we agreed that this approach might have more privacy concern and not quite fit our goal (although the idea of reaching out to the reviewers who are similar to you might be a nice feature to add to our solution).

Then we started constructing the flow model for our solution. We tested our flow with two scenarios we discussed before. 

Our flow model:

We decided that the next step to refine our solution would be defining the details for patients' profile (categories for lifestyle and personality) and situational values. We also started a document for final report.


Feedback from flowchart presentation:

• We need to keep in mind that users want to see the tool working as soon as possible, they will have very little patience for filling out detailed information about themselves (for their profile) up front. 

Need to identify the minimum threshold for required info about the user that is necessary for the tool to work properly. Aim for this threshold as the main use-case as opposed to trying to get all the information up front. Find a good way to prompt user for information after search results are presented --> "want better results? Fill out more information about yourself in your profile..."

• May be good to show example of the system prompting the user for more information throughout the interaction flow (when does this happen, what does it look like?)
• Want ability to start searching without logging in, but then have the option to save any inputed settings/preferences to their profile by signing in at the end of the search process.
• Would be good to show the parallel UX for the two main use cases: Someone who is using the tool as a guest (Have yet to sign into their premera account or fill out profile info), and someone who is signed in and has filled out their profile information.
• It will be important to show an example of a user profile, what does it include, what does it look like, where in the process would it be edited?
• Important to show results, how are they presented, how are they calculated?
• Focus on the user's ability to determine how they and a given reviewer are similar. Make it clear to the user how they are matched with a reviewer/review

Process Blog and Scenarios for Using the Tool (11/10 - 16)

(11/10 - 12 ) We started this week by preparing for Thursday's presentation on our concept. We met on Monday and Tuesday afternoons to develop our concept and finish the slide draft for feedback. After reading Linda and Karen's feedback, we decided to simplify our presentation of solution by just providing key information. We decided to present two slides with illustrations showing two important steps in our proposed solution.







(11/14 - 16) Chris went through all the feedback of our presentation, organized them in Google doc and we will discuss them next week. We met on Saturday to discuss possible scenarios for using the tool and we came up with four scenarios (each of us will make a sketch for each scenario):

1. A person with a painful toothache which also causes a headache, wants to see a dentist as soon as possible: This person is organized and likes his doctor to prepare a plan for him to follow situational attributes: urgency(availability, wait time), location

Result: The patient will type in the type of doctor he would like to see and would select urgent !! As a result the system will limit the amount of attributes for this patient and filter the results based on availability, wait time and location.

2. Plastic surgery: a person has enough time to find the right doctor. She would like to improve the look of her nose and she cares about her safety. She wants to make sure that her new nose looks better. She wants to connect with people who are the same age and have been through a similar process.

Profile attribute: Likes her doctor to make her feel safe and age

Result: The patient will type in the type of doctor. As a result the system provides more attributes to choose from (why does it provide more options?). The system will use the profile attributes to provide suitable situational attributes for this patient based on age. Doctors that have a good reputation for making their patients feel safe will have higher ratings..

3. Primary care doctor: a person don’t know exactly what type of doctor that he needs to see. He would like to meet with a primary care doctor who has good connections with other doctors so that he can point him to the right direction and recommend a good doctor to him.

Profile attributes: He likes his doctors to be talkative and caring

Result: The patient starts searching by selecting Primary care doctor. As a result, the system will provide this patient with relevant attributes to choose from that other patients with similar profile attributes have chosen before. In this case, the system will rank doctors based on their connections

4. Orthopedic Surgeon: A middle-aged patient is active but is considering getting a knee replacement because his knee is wearing down and preventing him from being active. This is not an urgent situation, and the patient has time to find the right doctor. They would also like to find someone who can help consult them to help decide whether the surgery is right for them. If it is, they want the doctor to help them create a recovery plan.

In their profile, this patient notes that they want to find a physician who is talkative and likes creating plans for patients. They also note that wait time is not much of an issue, they would rather find someone with a lot of experience.

When they go to search for orthopedic surgeons, the search tool provides a list of attributes that previous patients in similar situations have identified as important, such as: success rate, average recovery time, etc.

Physicians would then be returned in order of those receiving highest ratings for ability to create plans, helping patients reduce recovery time, success rates on this type of surgery, and experience doing knee replacements. Additionally those ratings would be produced mainly by people like the patient doing the search, so they know their are relevant and trustworthy.

Process Blog and Notes on Reviews (11/3-9)

Heuristic Evaluation Exercise

http://mhcid2014fall2ndprojectgjy.blogspot.com/2014/11/heuristic-evaluation-114-5.html

We narrowed down our concept to "search what people like me like". To refine our concept, we decided to do a little more research on reviews and explore our concept by sketching it out. The following is Chris' research note on reviews. We will meet on Monday to have consolidate our concept and prepare for Thursday's presentation.

http://www.sciencedirect.com/science/article/pii/S1567422312000464

• “word of mouth” reviews are critical for “service-based” products (like finding a doctor) because there is no way to test the service before you purchase.

• consumers associate this type of good as higher risk and uncertainty, therefore they put more weight on the information they can gather

• “when perceived risk is very high, consumers use word-of-mouth as their primary risk-coping strategy”.

• finding a doctor is considered a “credence service” and is considered the highest risk of all types of service goods because there is a high level of uncertainty.

• “[because of a greater level of uncertainty], consumers evaluating credence services are more likely to place greater emphasis on specific aspects of reviews”

• reviewer’s “reputation” is correlated with the perceived usefulness of a given review

• There is lots of uncertainty about the authenticity of reviews because of promotional incentives, don't trust motives of other reviewers

• “consumers must handle the uncertainty regarding the integrity and intentions of the people who write the reviews”

• peer ranking reviews as “most helpful” has been shown to be a good way to help with this. Unclear whether this method is relevant for healthcare implementations

• but it does provide motivation for the reviewer to provide a high quality review. Recognition is a very strong motivator.

• “experience and credence services are subject to variation in individual tastes and are therefore subjective in nature. Consumers are often highly confident about their own tastes and subjective evaluations but skeptical about the views of others”.

• “therefore, advice and comments from experienced customers helps review readers experience the service offering vicariously”.

• “perceived similarity between the message source and recipient is extremely important when evaluating credence services”

• “personal identification of the source based on information such as gender and geographical origin considerably enhances the credibility of the message”

• disclosing reviewer identity helps with credibility

• another study found this helps boost sales

• some studies have found that those who are shown to have more experience (or expertise) with a product or service gives their review more credibility

• but the study citing this found contradicting evidence...

• Some studies have found that consumers pay more attention to the textual portion of reviews than the star rating.

• the number of reviews associated with a product/service is often used as a heuristic to assess the general quality of the good.

• this should be incentive for Premera to generate more reviews

• Consumers trust peer reviews over “expert” reviews for experiential products.

• negative reviews are considered more helpful.

• important to show negative reviews

• Previous literature indicates that “consumers sometimes pick only reviews written by reviewers who, the consumers feel, are genuine and have similar social backgrounds, tastes and preferences. As a result, some reviews have greater influence on consumers than others”.

• although the main study citing this finding resulted in a different finding…

Heuristic Evaluation (11/4-5)

1) 3-5 desired experience and goals
2) Why is that important to the problem space
3) Example that do it well

1. Personalization
-created when consumers get the sense that the tool understands them, what they care about, and what their situation is, and thus the information is relevant to them. Users can trust that the tool understands their context, and also making the consumer feel that the tool is fully helping serve their needs.

Example:  Spotify

2. Privacy
As a tool integrates a more full understanding of a given user's context, privacy becomes a more relevant priority. A sense of privacy is achieved when consumers feel that information supplied to a tool is secure and won't be shared with anyone they don't want.

Example:  banking tools

3. Simplicity
The process of using the tool is straightforward such that it is easy to use for a wide range of users. The tool provides the right information at the right time.

Example: Amazon (one-click checkout)

4. Supportive
The tool goes beyond providing information by creating a support network or community. By doing this, the tool brings more emotion into the process.

Example: wearable health monitoring communities (fitbit communities for example). Specific facebook groups (those dedicated to rare diseases for example) can also provide support.

Brainstorming (10/30-31)

Multiple Group Brainstorm
We met with our pair group and began to brainstorm broad categories that interested us. We also tried using leading questions like "wouldn't it be great if..." to help generate problem areas and opportunities.

Single Group Brainstorm
We began our group brainstorming session by identifying 5 broad categories that we are interested in brainstorming concepts further":

• Scenario influence on presentation of information
• Integrating social network 
• Customize consumer experience based on their background/history
• Matchmaking 
• Reviews

We then brainstormed concepts/ideas individually to produce as many different ideas as possible (trying to reach the 50 benchmark). Once we felt we had generated enough concepts, we came back together and presented ideas for each broad category, placing sticky note concepts under each category.

Once all of our concepts were organized, we went back through each category and reviewed our ideas to identify themes across concepts. We wrote these basic themes in red to make it easy to consider concept areas we liked most. 

"Aha Moment": We liked the idea of having a "medical adviser", someone a patient could call, explain the situation, and they would provide individualized suggestions based on what they know about this. This idea was appealing because we felt patients would trust it much more than the current search tool. We then took a step back and realized that the "medical adviser" is really the role we would like the search tool to ultimately take, but what is missing is TRUST. We then identified trust as an overarching concept that we wanted our ultimate solution to integrate into a search tool. 

6 Major Concept Areas:

1) Category: designing how search information/data is presented based on the patient's situation

a. Start search by identifying situation, timeframe, etc. (generate search results)

b. Allow real-time re-ordering of search results based on patient's identified priority of care attributes (prioritizing search results)

2) Category: Integrating social network into search tool

a) Physicians in search results can have a badge indicating whether someone in your extended network (possibly via facebook) had used that doctor, provides way to reach out to them for more detailed feedback. (Irene)

b) Another way is to identify the situation and then generate connections in the social network from that, provide patient the option to connect with people.

3) Category: Reviews

Creating a review structure so that there becomes a consistent format of reviews to help consumers compare information, and also to ensure more rich information provided by reviews. Also would mitigate extreme reviews.

4) Category: reviews/data

Minimize information overload by only showing critical info up front, allow users to dig if they want to

5) Category: Reviews

Review input is abstract, paragraph style. System output (that consumers sees) is based on keywords

6) Category: Reviews (Irene)

Understanding reviewer background, compare to consumer preferences. Identify patterns in reviewer's review history. Prioritize reviewer's feedback based on shared values with the consumer. 

Pair Team Review (10-27)

Our group met with Acacio, Stewart, and Joy.

Going through their research findings was helpful to hear their thinking behind the project and see a different perspective on the information.

As we went through our frameworks we realized that we needed to break down our thinking behind each framework to make our presentation more clear and easy to follow. Joy had some suggestions on how to present our frameworks to emphasize the important information.

Influence of recent healthcare changes on searching for physicians

Guiding the next generation of public reports (see source)

• with insurance expansions under the Affordable Care Act, there has been an influx of newly insured people who now need to take control of their healthcare and make important decisions.
• additionally, previously insured consumers now covered under a new plan with a network of providers that does not include their previous providers not need to make decisions and would benefit from care information
• overall, the type and presentation of physician information is more critical than ever. 
• consumers now have more of a stake in the pricing of their healthcare plan, so being more transparent about the actual price information is important
• suggested gaps in information provided to consumers include:
• provider rates of success in helping patients maintain or restore health
• provider track record in ensuring continuity of patient care
• physician performance
• quality info that is paired with cost info
• is it possible to tailor the physician information to the individual consumer?
• look to these cited sources (Hibbard): 
• Best practices in public reporting NO.1
• Best practices in public reporting NO.2
• shaller consulting (2006)
• scan others for those that look interesting
• note keywords of this study

Supporting consumer healthcare decisions through data presentation (see source)

• recent changes in the healthcare system have made healthcare options more expensive but have given more control to consumers. Consumers are now more invested in their options and have increased motivation to making the right decision. This means that presenting choices in a way that helps consumers be better informed is increasingly crucial.
• an abundance of information does not necessarily lead to informed decisions
• current tools:
• include technical terms that are difficult for consumers to understand
• force consumers to compare multiple options on several variables
• require the consumer to differentially weight the options and factors

How do people shop online? What is the best way to report the information?

Word of Mouth Reviews in Hospitality and Tourism Management (see source)
-Discusses how hospitality services, like finding a doctor, is a distinct type of decision to finding and evaluating products.

• study asserts that for services that are intangible like hospitality, "interpersonal influence" is more important. This is the idea that consumers have a greater ability to influence eachother in decision-making.
• These types of decisions are high-risk

Information overload in e-commerce decision-making (see source)

• filters can help to alleviate perceived information overload
• individuals vary in their ability to process different levels of information
• more experienced online shoppers are able to process information more effectively and more efficiently

What makes a helpful online review (see source)

• for experience goods (those that require being experienced in order to be reviewed, like a doctor), extreme ratings (1 out of 5 stars, or 5 out of 5 stars) are less helpful than moderate ratings (3 out of 5 stars).
• taste plays a big role in ratings of experienced goods, therefore the majority of the ratings are extreme. However, consumers are often skeptical of the tastes of others, so these ratings aren't as helpful
• They find that lengthier, more detailed reviews are less helpful for experience goods versus "search goods" (things that don't need to be experienced to give a rating to, like cameras)

How to effectively present healthcare performance data to consumers (see source)

• the top 3 factors identified by consumers as most important are for quality of care:
• affordability of care
• doctor qualifications
• access to care for everyone
• too much information is counterproductive (even though consumers often report wanting more information)
• weighting or prioritizing different factors (of care) is difficult for consumers. Is this also true of physician factors??
• Currently, consumers must process large amounts of information, identify relevant factors and prioritize them, and bring all of this together into a choice. This is a difficult task for humans to do effectively. As a result, consumers aren't quick to use these tools, leading to low review/feedback rates (negative feedback loop). --> the problem
• proposed solution: make information more relevant to what consumers already understand and care about.
• avoid technical jargon, frame information clearly --> "physician is trusting and friendly"
• allowing users to drill down if they want can ease the burden of too much information up front. Make the specifics available but not in your face.
• make sure ratings metrics are consistent 
• summarizing and interpreting information for the consumer can be really helpful. This could mean visually organizing and ordering a list of physicians in a certain way
• use colors and symbols rather than numbers when possible
• The problem with cost information: Americans are biased to believe that higher cost indicates higher quality, which is not always the case. Because of this, cost information should be shown within quality strata.

Supporting consumer healthcare decisions through data presentation (see source)

• consumers tend to take shortcuts when there is an informational overload in the decision process. 
• overweight the factors they understand
• overweight one factor over other important factors to reduce the number of choices
• most current tools that help consumers compare options are based on the assumption that consumers know what is most important to them and can therefore weight the presented information effectively to make their choices. This has been shown to be an inaccurate assumption.
• when people are in a situation that is both complex and unfamiliar, they likely do not have fixed ideas about what is important to them. In this case, information presentation has a significant impact on what is attended to and used.
• even when information is understandable to a consumer at an analytic level (knowing what the numbers mean), it may not be understandable at an emotional level (how good is a 4/5 rating?).
• use symbols/colors instead of or in addition to numbers alone

Process Blog by 10/26/2014

This week we continued our secondary research by focusing on the following questions:

• What do patients care about when they search for physicians?

• What distinct categories of patients are there? How do their values differ when searching for doctors? 

• How do physicians interact with current tools, what do they care about?

• What are the shortcomings/issues associated with current physician-search tools?

• What are the main attributes that Premera competitors are focusing on? 

• How have the recent changes in the healthcare system altered who searches for doctors, what they care about, etc.?

• Research the way people shop online? What rating systems/metrics are most effective? Is there research that compares the way people shop online for products versus doctors/people/services?   

We recorded our finding for each topic in separate blog posts.

To construct our research framework, we first wrote down our research topics and findings on sticky notes, and organized them on the white board. 

After we talked through all of our findings, we picked out two research topics we found extremely exciting to construct our two research framework. 

We will meet with another group next Monday to have them review our frameworks and revised problem/ opportunity/ hypothesis.

How do physicians interact with current tools, what do they care about?

Stimulating the Adoption of Health Information Technology
http://www.nejm.org/doi/full/10.1056/nejmp0901592

Congress apparently sees HIT — computers, software, Internet connection, telemedicine — not as an end in itself but as a means of improving the quality of health care, the health of populations, and the efficiency of health care systems.

The Impact of the Internet and Its Implications for Health Care Providers: Findings From the First Health Information National Trends Survey
http://archinte.jamanetwork.com/article.aspx?articleid=766849

As a result, data entry and review systems are becoming ubiquitous in hospital settings and health care technologists are crafting applications in medical informatics to improve the quality of health care.

The implication for physicians is that patients may come into the office with printouts that they did not personally collect; the implication for Web site designers is that health information seekers may be looking for easy-to-print materials to give others.

Those who searched online the most tended to be younger, women, and better educated, and tended to have higher incomes than those who searched the least.

What Does It Cost Physician Practices To Interact With Health Insurance Plans?
http://content.healthaffairs.org/content/28/4/w533.full

An approximate calculation can be made by multiplying the number of physicians in U.S. office-based practice by the $68,274 per physician annual cost we found.

Health plans claim to have taken steps to reduce the administrative burden they place on physicians; our data suggest that physicians continue to be discontented with this burden.

The average physician reports spending nearly three weeks per year on these interactions (nearly two weeks if the median value is used); twenty-three weeks per physician per year are spent by nursing staff, and forty-four weeks, by clerical staff.

A small fraction of this time—two hours per year by the average physician, and little more by their staff—is devoted to providing quality data to health plans or to reviewing quality data generated by plans about the physicians’ practice. Primary care physicians, especially those in small practices, spend larger amounts of time interacting with health plans than physicians in other specialties.

What distinct categories of patients are there? How do their values differ when searching for doctors?

Peer-to-Peer Health

http://www.pewinternet.org/2011/02/28/peer-to-peer-health-care-2/

When finding recommendation for a doctor or a specialty, 52% of adult say they think health professionals are more helpful than peer sources when it comes to getting a recommendation for a doctor or specialist, 27% of adults prefer to ask friends family and fellow patients for this type of advice. Order adults are much more likely than younger ones to say health professionals’ opinions are more helpful. Generation X internet users (34-45) are the most likely age group to look online for information about health professionals.

When finding recommendation for a hospital or medical facilitates, 62% of adult say they think health professionals are more helpful than peer sources when it comes to getting a recommendation for a doctor or specialist, 27% of adults prefer to ask friends family and fellow patients for this type of advice. Older adults are considerably more likely than younger adults to turn to health professionals for this type of information

Online Health Seeking: How Social Networks Can Be Healing Communities

http://www.pewinternet.org/2010/10/25/online-health-seeking-how-social-networks-can-be-healing-communities/

1.  Internet

Adult: 61% get health info online (80% are internet users)

E-Patients: 60% consume social consume social media; 29% have contributed content; 19% consult rankings of providers (5% post them); 18% consult reviews of hospitals (4% post them)

2.  Other Data

60% of e-patients say the info found online affected a decision about how to treat an illness or condition.

53% say it lead them to ask a doctor new questions, or to get a second opinion from another doctor.

38% say it affected a decision about whether to see a doctor.

On Going to See the Doctor, the Contributions of the Patient to the Decision to Seek Medical Aid: A Selective Review
http://www.sciencedirect.com/science/article/pii/0021968163900997 

Features
1. Most participation and utilization studies, while seemingly similar to the initial process of seeking medical advice, are really the repeated use or reuse of the doctor-patient relationship. 

2. It is the choice between types of medical service, between public or private care, general practitioner or specialist, clinic group or solo practice, Doctor X or Doctor Y-choices in medical service about which there is great debate on what is good, better or best.

Factors influencing the decision to seek medical aid: the perceptions of and beliefs about symptoms, the expectations of medical personnel, and the rationale of the need for medical care. 

The decision is not necessarily idiosyncratic but often rooted in the patient’s social background.

Complexity, Public Reporting, and Choice of Doctors: A Look Inside the Blackest Box of Consumer Behavior
http://www.ncbi.nlm.nih.gov/pubmed/23999489

The quality of choice erodes dramatically as choice sets incorporate more options and more performance metrics, even though the information on SelectMD was far simpler than on websites that also include cost measures, safety metrics, and assessments of electronic access.

Second, certain types of consumers are more susceptible than others to the challenges of complexity. Some of these susceptibilities relate to decision skills, others to decision styles.

Third, incorporating patient comments into public reporting clearly displaces consumers’ attention from standardized performance metrics, leading to choices that are suboptimal on those measures.

Use of  filters helps relieve cognitive burden by reducing the size of a choice set to one that seems manageable. In so doing, consumers often exclude the best clinicians 

What are the shortcomings/issues associated with current physician-search tools? What are the main attributes that Permira competitors are focusing on?

Search attribute	ZocDoc	UPMC	WebMD	Vitals	DoctorDirectory
Specialty	yes	yes	yes	yes	yes
Zip code	yes	yes	yes	yes	yes
Insurance	yes			yes
Doctor name		yes	yes	yes
Doctor Condition			yes	yes
Procedure			yes

Search Results:

http://www.zocdoc.com/

http://findadoc.upmc.com/FindADocSearch.aspx

http://doctor.webmd.com/

http://www.healthgrades.com/landing/dcp/sea-20140910?s_tnt=184566:5:0

http://www.doctordirectory.com/findadoctor/

http://www.vitals.com/

The pros are you can make an appointment with a Doctor online, and easy. The cons not many Doctors use Zoc Doc service. So you will have a very shallow network of Doctors to go to.

http://www.quora.com/What-are-the-pros-and-cons-of-using-ZocDoc-to-make-doctors-appointments