Reduce clinical contact center workflow costs with AWS

Amazon Connect along with its ecosystem of cloud services helps you reduce your contact center costs tremendously. Here’s a short guide on how we achieve economies of scale for our customers.

Estimated reading time: 17 minutes

    Email article to yourself or a colleague to read later.


    Running a clinical contact center is not cheap. Healthcare leaders know that.

    It’s fast moving – much like every other contact center but is burdened heavily with HIPAA, PHI and other severe restrictions.

    Our clinical contact center team lives, breathes and thrives in this environment.

    Amazon Connect along with its ecosystem of cloud services helps you reduce your contact center costs tremendously. Here’s a short guide on how we achieve economies of scale for our customers. 

    Feel free to use this guide within your own contact center or contact us to get it done for you.

    Healthcare workflow – where do you spend time and money?

    • Contact center
    • Revenue cycle management 
    • Print
    • Scan
    • Fax
    • Phone tag
    • Data entry
    • Paper.. paper…

    Let’s just agree on the above. Healthcare workflows are not rocket science, but are notoriously labor-intensive.

    Let’s look at each component next.

    Clinical contact center cost components

    If you work at a clinical contact center, you know that a large part of your day goes into handling the following:

    “Answering service” for pharmacies, patients
    Appointment scheduling – calls, emails
    Appointment reminders
    Reappointing no-show patients
    Reappointing cancelled patients
    Recalling patients fallen out of care
    Website Chat & Social Media responses 
    Fax – monitoring, responses
    Community Outreach Service
    Surgery Scheduling Services
    Outreach for updated insurance information
    Prior-authorizations
    For certain payers, get PCP referrals as needed
    Outreach to patients for demographics updates
    Patient Pre-registration / intake
    Patient COVID screening 
    Patient surveys
    Patient review requests
    Patient text messages / 2 way conversations 
    Patient message blasts – practice updates 
    Physician Referrals (inbound)
    Physician Referrals Marketing
    Verify coverage on primary, tertiary, Medicaid etc.
    Discharge follow up calls
    Care Gap Outreach
    Eligibility calls (specific specialties)
    Wellness / end of year patient outreach
    Denials payer conversations
    Clinical contact center cost components

    Some of these duties do, technically, bleed into the revenue cycle management domain. Sure. We understand that. 

    But, we still have to get these done. There’s always a front end and a backend part of the revenue cycle management teams. You are that “front end” part of the revenue cycle management team.

    Answering service

    We spend a lot of time answering inbound calls from new and existing patients, referring provider offices, pharmacies, vendors, spam / robo calls, sales calls etc. And on top of that, these calls are in various languages – languages that our patients are most comfortable speaking in. We need to support all those.

    Not all of us have the luxury of supporting all languages that our patients speak – even if we want to. Some of our hospitals with large budgets tend to have translation services teams to handle such calls but many of our smaller hospitals do a shoddy job with these non-standard calls.

    More often than not, website visitors go unattended. Most hospitals do not have a website chat service available – no humans available to answer website chats.

    Most hospitals do not even have social media chats available. 

    Each day, each minute – hospitals lost potential patients. And are not even aware of it.

    Our contact center teams live and breathe multi channel support for our healthcare customers day in and day out. We face this problem all day long. We alleviate our challenges at scale using Amazon cloud services. 

    You can do the same as well – here’s the combination of AWS services we recommend you use to solve your multi channel or even omni channel patient engagement / patient access challenges in today’s connected world.

    • Use Amazon Connect
    • Introduce IVR
    • Introduce self service
    • Use Amazon Chime for technicians
    • Integrate with Pinpoint for SMS
    • Use Amazon Lex and Polly

    Again, of course, you could also simply use NisosHealthCRM.

      Email article to yourself or a colleague to read later.


      Appointment scheduling calls

      Many of our inbound calls are related to appointments. Many of our outbound calls are related to appointments. Whether they are inbound or outbound calls, we always have to dispose of the call with what the call was about, take notes on what transpired during the call. 

      This is the dreaded ACW (after call work). A waste of time.

      This, then falls into the territory of Quality Assurance staff that have to manually check, validate that call quality is being adhered to, all quality check marks are being “checked off”. The Ts are crossed and the Is are dotted.

      The calls need to be recorded and the private health information needs to be redacted.

      Sometimes the inbound calls are related to payments and instead of directing the patient to pay their bill via check (in other words, potentially losing the chance of collecting a sure-shot revenue event) we need to accept payment from the patient in a secure manner while not storing any of the patient payment information.

      The inbound caller always has to identify themselves – even when they might already be an existing patient of the practice for the last 10 years.

      The inbound caller might have to wait for a LONG time on hold to speak to the front desk even if all they want to know is the location of your office or even just their own appointment time.

      Or to even just reschedule their appointment.

      There are so many such opportunities of improvement.

      Endless.

      For inbound calls, our recommendation is to use a combination of these AWS Services

      • Use Amazon Connect
      • Introduce IVR
      • Use Everyone API
      • Use Amazon Contact Lens
      • Introduce self service
      • Use HIPAA secure EMR data dips for maximum self service 
      • Integrate with Pinpoint for SMS
      • Integrate with chatbots
      • Use Lex and Polly

      Again, of course, you could also simply use NisosHealthCRM.

      Re Appointment calls

      Oh, the joys of making outbound calls to fill up the appointment calendar of your hospital 🙂 

      There’s the list of no shows that you need to reappoint.

      Then you have the list of patients that have fallen out of care – that you need to call and ensure that you see them every quarter or 6 months or whatever you have.

      You also have your list of cancelled appointments that you can call to reappoint – patients don’t always make the best decisions. It is always best to reach out to them and make sure they see your provider for their initial chief complaint, even though the original symptom might have receded. You don’t want that to blow up to an ER visit that increases the utilization rates.

      Of course, with outbound calls, you always have the challenge of “voicemails”.

      Nowadays, with the advent of wonder features like “silence unknown callers”, most of your calls are going to “voicemails” anyways.

      So, you are going to have to start relying on voicemails more and more. Or, you are going to have to start relying on sending text messages more and more.

      Even with text messaging, these days, phone companies have introduced the “great” “filter unknown senders” features – your text messages are going into the so called “unknown sender” filter.

      Butk what can we do? We have to play in that world.

      Our recommendation is to use a combination of these AWS services to reduce the workload:

      • Use Amazon Lambda
      • Use HIPAA secure EMR data dips
      • Use Pinpoint for SMS and email
      • Use “request call” logic
      • Use Amazon Connect
      • Integrate with chatbots
      • Use Amazon Lex and Polly

      Again, of course, you could also simply use NisosHealthCRM.

      Website chat, social media responses

      This is an interesting one because every single patient or website visitor expects you to help them when they are visiting your hospital website.

      Meanwhile, every single one of them expects you to respond on various social media channels as well – be it facebook, google, instagram, twitter etc.

      But who has the budget or the manpower to staff all these channels anyway? Same goes for our healthcare customers – they don’t have dedicated teams monitoring these channels and at the same time, they don’t want to be unavailable on these channels either. 

      The same concept goes with referral management as well – accept inbound referrals from multiple referral sources/channels, but consolidate them all into one software/channel that you manage/monitor so there are no leakage on leads and no lack of communication(s).

      Our recommendation is to consolidate all your communication channels into one – Amazon Connect Web Chat. 

      This way, your asynchronous chat communications can all be consolidated into one channel – chat. This chat can have messages sent from various social media sources to Amazon Connect Web Chat. 

      Amazon Connect Web Chat already integrates very well with Amazon Connect Contact Center, therefore the same agents that are answering your inbound and outbound calls can handle these chats. 

      What’s more – you can use Lex and Polly to converse with and in any languages that Amazon supports, with patients speaking any language!

      Now you have all your multi channel bases covered.

      • Use Amazon Lambda
      • Combine response channels into Amazon connect
      • Use Amazon Lex and Polly

      Again, of course, you could also simply use NisosHealthCRM.

      Patient intake

      Patient intake is where you can save a LOT of time – we save an average of 25-30 mins per patient overall. Our patients finish almost everything before they even step into the practice.

      Take a step back and think about the non billable activities that really eat up your staff time (and why you need to hire more staff).

      Your patient comes in and needs to be checked in. They need to hand over their insurance papers that need to be scanned into the EMR. They fill out the patient intake form that then needs to be typed in (to some extent) into the EMR and the rest needs to be scanned into the EMR. The consent forms need to be scanned into the EMR.

      Next, the patient pays the copay and gets a receipt that’s printed out and then the patient waits.

      After the technician takes the patient into the room, they spend a good 10-12 mins “working up the patient”. In other words, chief complaint, history of present illness, medical history, surgical history, social history, medication history etc.

      Then, the vitals and if there were any charts or scans that were relevant, the patient is asked to “remember to send them” to the doctor for analysis. If the patient is not in possession of the scans, then the patient is asked to sign a HIPAA consent to release medical records form so that your provider can request these records from the patient’s provider. 

      Then this fax is sent to the patient’s provider and the waiting game begins.

      All these things, other than the patient’s vitals could have been done by the patient from their homes.

      Including pre-paying for the appointment – which would have doubly ensured that the patient would have avoided being a no-show.

      We do that. We ensure that our patients complete all their intake by themselves and if they do not, an alert goes out to our patient coordinators and they call the patient before the appointment to complete the intake process before the patient comes in.

      We recommend that you use a combination of these AWS services to achieve economies of scale with the above.

      • Use Amazon Lambda
      • Use Amazon Pinpoint campaigns
      • Use HIPAA secure EMR data dips for patient intake 
      • Use Lex and Polly
      • Use S3 for patient docs
      • Use Textract to extract text from images/docs

      Again, of course, you could also simply use NisosHealthCRM.

      Patient screening, surveys, reviews, SMS 

      These days we all screen patients for COVID before they come in for their appointments. This involves calling them and asking them a series of questions. That’s about it. 

      If you are a good hospital operator, each quarter, you are probably running patient and provider satisfaction surveys. If you are one of the over 4,000 hospitals that participate in HCAHPS surveys, you know that CMS is running surveys with over 3.0 million patients that complete the survey each year. This survey runs between 2-42 days after each hospital stay.

      Hopefully you are a true believer in google reviews and are collecting reviews each day within 15 mins of the patient leaving your offices. There are many ways of asking for reviews – via text messages, email or by calling the patients when you do not have their SMS or email contact information.

      All of this really adds up to the workload of the patient contact center staff and while these are absolutely crucial in their own ways, these always fall “to the wayside” because they are not considered crucial for “clinical reasons”. We recommend you automate most of these using AWS services as below:

      • Use Amazon Pinpoint campaigns
      • Use HIPAA secure EMR data dips 
      • Use Lex and Polly
      • Use S3 for patient docs
      • Use Textract to extract text from images/docs

      Again, of course, you could also simply use NisosHealthCRM.

      Processing incoming faxes

      Ahh.. don’t we all love faxes? 🙂 We have to live and breathe faxes each day. 

      While most healthcare IT folks tell us that faxes are a dying technology, our medical billing and call center folks will tell you otherwise.

      Our healthcare IT team actually recommends that you fall in love with your fax machine all over again… using Amazon Web Services.

      Faxes are the center of the healthcare universe, and you need to learn to learn to love them. 

      Whether you are handling incoming faxes for referrals, EOBs, ERAs, Medical record retrieval requests from larger groups or smaller providers, prior authorizations, eligibility responses, TCDs, medical clearances, spam faxes, vendor marketing faxes, prescription related faxes, EPAs etc. you need to handle them.

      Faxes do eat up a lot of time.

      You need to type and transfer information from the fax to the EMR or PMS – wherever it was intended to go to. You need to print, scan, upload to the EMR or PMS – wherever you need the information to go to. You need to keep track of these faxes (file storage). And this list goes on.

      Optical character reader technology has been around for ages but it has never been the best for reading images.

      Faxes are not the easiest to read since they are, technically, images.

      Amazon Textract does an excellent job at reading images and extracting text information from faxes. Our entire revenue cycle management system and revenue cycle management team is built around this – Amazon Textract. 

      We recommend you use a slew of AWS services to scale your team as well.

      • Use Amazon Chime  T38 fax / Hylafax
      • Use S3 for patient docs
      • Use Textract to extract text from fax images/docs
      • Use Lambda
      • Use AWS step functions
      • Use DynamoDB for documents database 

      Again, of course, you could also simply use NisosHealthCRM.

      Amazon Textract will allow Human intervention- don’t worry!

      Physician referral marketing

      Referral marketing – not many people do it well. They don’t know the value of physician referrals well enough and certainly do not know how to market to referral physicians around their neighborhood. 

      We have seen physician group leaders hiring staff and expecting them to just go “door to door”, expecting them to bring referrals pouring in. They do not equip their staff with the right tools or technologies to do a good job with.

      Done professionally, physician referrals are a golden opportunity to grow your healthcare business. Sure, the NPPES database is not the most accurate when it comes to physician locations or associations with various practices but we have noticed that phone numbers and fax numbers are fairly accurate.

      If you use faxes and phone calls to market your hospital to referring physicians, you will grow a tight and high performant network of referring physicians.

      You don’t actually need to spend a whole lot of time or money on managing some huge database either. With Amazon’s S3 and Athena, you can get by with just CSV files.

      In addition to that, you can use Amazon Connect and Amazon Pinpoint to run your campaigns in auto mode 🙂

      Here’s the list of AWS services we recommend you use to do marketing at scale.

      • Use NPPES database (weekly updates)
      • Use Amazon Athena
      • Use DynamoDB or RDS
      • Use Amazon Chime T38 or Hylafax 
      • Use Amazon Connect
      • Use Amazon pinpoint campaigns

      Again, of course, you could also simply use NisosHealthCRM.

      Community outreach marketing

      We haven’t seen very many hospitals doing community outreach as it is a bit expensive. However, we have seen good results with this approach. We purchase consumer data for our healthcare customers – around their practice locations and use our tried and tested set of AWS services to run marketing campaigns.

      Most calls do seem to go to voicemails but these are expected as we are only introducing our customers’ businesses to the neighborhood – not asking for immediate business.

      Yes, each day, we do end up getting new appointments as well. But that conversion rate is almost never more than 4-5%.

      Here are the recommendations:

      • Use payer assignments
      • Or use DataAxle CSVs
      • Use Amazon Athena
      • Use DynamoDB or RDS
      • Use Amazon Connect
      • Use Amazon pinpoint campaigns

      Again, of course, you could also simply use NisosHealthCRM.

      Prior authorizations 

      You have emergency prior authorizations that come up in the middle of the day 🙂 Yup, those happen. The patient is in the office and needs a procedure done. Sure, you can schedule the patient for a later date but the patient might be a no-show.

      You know based on your experience there’s a very good chance that if you submit the prior authorization request you might get it approved within an hour.

      We have written about how to automate part of your prior authorization workflow here. Use a combination of open source software and cloud Amazon Web Services combination as stated below. 

      • Use Hylafax or Amazon Chime T38
      • Use Amazon Lambda
      • Use HIPAA secure EMR data dips
      • Use DynamoDB or RDS
      • Use Amazon Textract
      • Use Amazon Connect

      Again, of course, you could also simply use NisosHealthCRM.

      Eligibility verification

      One of the BEST ways to handle patient eligibility and to avoid insurance eligibility issues is to do real time eligibility checks. There’s a trick to this.

      When the patient wants an appointment, you give them an appointment date/time, ask for their insurance detail and run their eligibility right then and there – that’s it. It gives you their eligibility for that moment.

      This does not guarantee that they will be eligible by the time they come in for their appointment though – they might get laid off or quit their job or terminate their coverage before they come in for their appointment, but at least, you are not blocking the appointment for someone that is not “covered” in the first place.

      If you believe that denials is the heart of an RCM team, then you will soon discover that eligibility and insurance verifications are something that you can control very well – especially since these are easy to get done and very easy to address by reaching out to patients ahead of their appointments.

      Whether you use your practice management system’s eligibility system or not, these are crucial to get done. We will also show you how to get these done in bulk using a simple spreadsheet as well (although, it is a bit geeky, to be honest).

      Otherwise, simply use one of the best APIs in the market – changehealthcare APIs or Navinet’s APIs. These guys are doing a stellar job of providing API and portal based access to eligibility and insurance verifications (ad hoc / single transaction and batch ones).

      The responses you get from your eligibility requests on these vendor sites will vary, and hopefully you understand (by now), how to decipher the response received from payers w.r.t eligibility requests.

      If you are going to use AWS cloud services, here’s the combination we would request. If you are not going to use changehealthcare APIs, you can also connect to each payer’s EDI API endpoint individually as well.

      • Use Hylafax or Amazon Chime T38
      • Use Amazon Lambda
      • Use HIPAA secure EMR data dips
      • Use DynamoDB or RDS
      • Use Amazon Textract
      • Use Amazon Connect

      Get necessary PCP referrals automatically 

      This one falls through the cracks ALL the time and almost always ends up in a denial downstream.

      Your patient eligibility runs through well, but there’s a small note that your team misses – this patient needs a referral to see your provider. 

      That’s deadly – because without this referral, you are not going to get reimbursed.

      As soon as you run your eligibility you should get an alert that you need to get a PCP referral. This should trigger a fax to the PCP that’s listed on your patient’s profile. Then, your system should wait for the faxes coming in and as soon as a fax comes in that’s a “needle in the haystack” and is the referral fax of the patient’s PCP referral, it should alert you that the referral has been received.

      Meanwhile, if the referral has not been received 2-3 days before the patient’s appointment, an automated call should go out to the patient asking them to reschedule the appointment because the “clearance referral” has not been received. This should allow the patient to “press 1 to talk to a scheduler” so that they can reschedule their appointment OR they should be able to pick the next available appointment on their own.

      All of this is possible via a combination of these aws services.

      • Use Hylafax or Amazon Chime T38
      • Use Amazon Lambda
      • Use HIPAA secure EMR data dips
      • Use DynamoDB or RDS
      • Use Amazon Textract
      • Use Amazon Connect

      Again, of course, you could also simply use NisosHealthCRM.

      Wellness campaigns, end of year campaigns, discharge follow ups, care gap follow ups, other outreaches.

      Each month and each quarter.. Plus end of the year, there are massive campaigns that your contact center team needs to run.

      Each month you could be running various campaigns for wellness e.g. (CDC) COMPREHENSIVE DIABETES CARE , (CBP) CONTROLLING HIGH BLOOD PRESSURE

      (COA) CARE FOR OLDER ADULTS , (COL) COLORECTAL CANCER SCREENING, (BCS) BREAST CANCER SCREENING , (CCS) CERVICAL CANCER SCREENING , (CIS) CHILDHOOD IMMUNIZATION STATUS, (IMA) IMMUNIZATIONS FOR ADOLESCENTS , (W15/W34/AWC) WELL-CHILD AND ADOLESCENT WELL-CARE VISITS , (ABA) ADULT BMI ASSESSMENT etc. 

      Or, you could be running discharge follow up campaigns each day based on discharge dates of patients.

      Care gap follow up campaigns are run every single day to bring patients back to the appointment calendars.

      There are several outreaches that are run on a daily basis that are of massive scale. 

      There’s no way to scale humans to this level and you need to rely on automation in your contact center – with text messaging, Whatsapp, voice mails, voice calls/push to talk, etc.

      These are the AWS Services that we use and also recommend that you use to scale your campaigns:

      • Use Amazon Pinpoint 
      • Use Amazon Lambda
      • Use HIPAA secure EMR data dips
      • Use Amazon Connect

      Again, of course, you could also simply use NisosHealthCRM.