Diabetes cycles of care incentive payments, Kensington 2019

Incentive payments

Current practice diabetic patient population at Kensington site : 256 (according to early 2018 PIP data) Annual outcome payment : $20 per diabetic patient = $5120. : has been achieved in recent years, but not in the most recent quarter.

Current Diabetes SIP claiming rate (according to late 2018 PIP data) = 49%

Target Diabetes SIP claim rate to claim annual outcome payement = 50% target has been achieved in recent years, but fell below 50% in the most recent quarter

Additional Diabetes SIP revenue, if target Diabetes SIP claim rate achieved = \(\frac{50-49}{100} \times 256 \times \$40 = \$100\)

Additional revenue if Diabetes Outcome Payment achieved = $5120 (annual outcome payment) + $100 (additional diabetes SIP revenue) = $5220

Tasks required to improve detection of patients who might benefit from diabetes review

In order to improve the care of patients with diabetes, it is necessary to identify patients who have diabetes. This is in part achieved by ensuring that patients with a diagnosis of diabetes in the past history have a coded diagnosis of diabetes in the file.

This is because if the diagnosis is not coded then automated techniques to identify patients with diabetes do not work well.

For example, the Best Practice clinical software does not recognize diabetes diagnoses of the type ‘non-insulin dependent’ or ‘insulin dependent’ diabetes or the variations of that style of diagnosis naming (e.g. ‘NIDDM’ and ‘IDDM’).

SQL code at the end of this document is an example of technique to help identify those patients in the file with the ‘NIDDM/IDDM’ style of diabetes diagnosis recorded in file, without any other diagnosis which is recognized by Best Practice as ‘diabetes’.

That search identified 128 patients across coHealth with ‘iddm’ or ‘diabetes’ in a recorded diagnosis, but not identified as diabetic by Best Practice. Of these, twenty-four (24) had a ‘iddm’/‘niddm’ style diagnosis.

Using this, and other, techniques a health care provider (in this case, Dr Kayty Plastow) can then amend the clinical record so that patients with a diagnosis of diabetes is then recognized as being ‘diabetic’ in subsequent searches using Best Practice.

Daily detection of patients with diabetes who already have an appointment to see a health care provider

More than 50% of eligible patients each week seen at the practice left without having a recorded completion of diabetic cycle of care.

In other cases, the cycle of compare might be complete, or be almost complete, but the attending health provider might not be aware that the cycle is close to completion.

In another project at Kensington site, patients were identified every day for a period of six weeks, who were already booked to be seen that day. Clinicians were notified that a preventative action (immunization with Zostavax to reduce the risk of shingle) could be done. During the intervention period, immunization rates for Zostavax improved in absolute terms by 8.9%, a 17% relative increase in Zostavax immunization in the target population.

It was planned that diabetic patients who have not had diabetes SIP claimed be identified daily, and clinicians notified.

Patients booked in each day during the intervention period (20th March to 30th April 2019) were identified for potential cycle of care completion eligibility. There are many reasons why this might the case. Completing a cycle of care takes time, and also takes patient motivation to complete some elements of care. Given the presence of competing issues, health care providers might simply need to be reminded that preventative and monitoring health activities are potentially overdue.

A Best Practice SQL script which identifies non-immunized eligible patients is presented at the end of this document. Eligible patients were identified at the beginning of each day for a month, and reminders sent to clinicians (nurses and doctors) that these patients were eligible.

Modified variations of these tools can be used to identify patients for whom other health screening activities are recommended, and both measure and monitor practice performance in these activities. Examples include bowel cancer or cervical cancer screening, or early detection of chronic kidney disease among patients with diabetes.

Plots

Regression analysis

Linear regression analysis, with interaction effects, shown in table below.

The ‘discontinuity’ of the Kensington proportion of active diabetic patients with a completed diabetes cycle of care at ‘day zero’ (2019-04-09) is significant, with a jump of 14% (standard error 1.62%, p-value 1.16e-05).

Following the treatment period, the rate of Diabetes SIP claims among Kensington patients was inferior to prior to the intervention .

Impact

During the treatment period itself, an estimated 14.0% (95% confidence interval 10.9% to 17.2%) absolute increase in reported completion in diabetes cycle of care among 275 active patients with diabetes, about 38 extra patients.

Increase in revenue due to Diabetes SIP payments was \(\$40\times38 = \$1520\). If, as very likely, the government-set target of 50% Diabetes SIP coverage was achieved, then there was an additional \(\$5520\) revenue per quarter. The estimated total increase in revenue is \(\$7740\).

Each completed diabetes cycle of care is an opportunity for additional interventions which reduce or manage diabetic complications (such as reduction in blood pressure, sugar levels or cholesterol). If half of the additional 38 patients who had a completed cycle of care also had an additional effective intervention, then there could be four or five less heart attacks in the practice diabetic population over five years.

There is potential for additional benefit after repeated treatments. Methods used in repeated treatments may need further refinement to better target the ‘unreached’ target population.

Detection of patients who are likely to have diabetes, but have no ‘coded’ diagnosis

SELECT *

FROM BPS_Patients
  WHERE StatusText = 'Active'
  AND (InternalID IN (SELECT InternalID FROM PastHistory WHERE ItemText LIKE '%iddm%')
  OR InternalID IN (SELECT InternalID FROM PastHistory WHERE ItemText LIKE '%diabetes%')
)

AND InternalID NOT IN (SELECT InternalID
  FROM PastHistory
    WHERE ItemCode IN (3, 775, 776, 778, 774, 780, 1563, 7840, 11998)
    AND RecordStatus = 1)

AND InternalID NOT IN (SELECT InternalID
  FROM PastHistory
    WHERE Itemtext LIKE '%pre diabetes%'
    AND RecordStatus = 1)

AND InternalID NOT IN (SELECT InternalID
  FROM PastHistory
    WHERE Itemtext LIKE '%prediabetes%'
    AND RecordStatus = 1)

ORDER BY surname, firstname

Number of ‘active’ diabetic patients who have a SIP claimed

This R code generates a SQL scripts for listing active diabetic patients who have a Diabetes SIP claimed. A modified versio nof the generated SQL script lists all active diabetic patients, with or without Diabetes SIP claims.

sqlK.query <- 
  "SELECT *
FROM BPS_Patients

WHERE StatusText = 'Active'

AND InternalID IN (SELECT InternalID 
    FROM Visits v
    INNER JOIN (VALUES('%%bhagwat%%'),('%%fong%%'),('%%ekanayake%%'),('%%shoesmith%%'), 
            ('%%plastow%%'),('%%samarawickrama%%'),('%%obeyesekere%%'),('%%chaves%%'),
            ('%%ryan%%'),('%%mikhail%%'),('%%haynes%%'),('%%buckwell%%'),('%%maxwell%%'),
            ('%%grace ho%%'),('%%bullen%%'),('%%lambrou%%'),('%%zeigler%%'))
        AS ProviderName(Name)
        ON v.DrName LIKE ProviderName.Name
    WHERE VisitDate BETWEEN DateAdd(Year,-2,'%s') AND '%s'
    AND RecordStatus = 1

    GROUP BY internalid
    HAVING count(internalid) >= 3)

AND InternalID IN (SELECT InternalID
    -- history of diabetes
    FROM PastHistory
        WHERE ItemCode IN (3, 775, 776, 778, 774, 7840, 11998)
        AND RecordStatus = 1)

AND InternalID IN (SELECT InternalID
    FROM Invoices WHERE InvoiceID IN (SELECT InvoiceID
        FROM Services
            WHERE Recordstatus = 1
            AND Servicedate > DateAdd(Month, -6, '%s'))) 

AND InternalID IN (SELECT InternalID 
    # remove this section if finding 'total' diabetes patients
    # with or without SIP
    FROM Invoices WHERE InvoiceID IN (SELECT InvoiceID 
        FROM Services WHERE Recordstatus = 1 
            AND Mbsitem IN (2517, 2521, 2525) 
            AND Servicedate BETWEEN DateAdd(Year, -1, '%s') AND '%s'))

ORDER BY surname, firstname
"

sql.query <- 
"SELECT *
FROM BPS_Patients

WHERE StatusText = 'Active'

AND InternalID IN (SELECT InternalID 
    FROM Visits v
  INNER JOIN (VALUES('%%membrey%%'),('%%halewood%%'),('%%maccartney%%'),('%%coles%%'),('%%dowd%%'),
    ('%%winter%%'),('%%wells%%'),('%%wiener%%'),('%%rainsford%%'),('%%tomlinson%%'),('%%walker%%'),
            ('%%linton%%'),('%%victoria%%'),('%%gordon%%'),('%%johnston%%'),('%%mascarenhas%%'),('%%khaira%%'),
            ('%%green%%'),('%%wilding%%'),('%%mancey-jones%%'),('%%gardiner%%'),('%%pike%%'),('%%scopel%%'),
            ('%%ohnmar%%'),('%%rengasamy%%'),('%%taylor%%'),('%%leung%%'),('%%guo%%'),('%%bui%%'),
            ('%%whiting%%'),('%%keaney%%'),('%%bourke%%'),('%%hewett%%'),('%%anderson%%'),('%%whiting%%'),
            ('%%bramwell%%'),('%%khaira%%'))

        AS ProviderName(Name)
        ON v.DrName LIKE ProviderName.Name
    WHERE VisitDate BETWEEN DateAdd(Year,-2,'%s') AND '%s'
    AND RecordStatus = 1

    GROUP BY internalid
    HAVING count(internalid) >= 3)

AND InternalID IN (SELECT InternalID
    -- history of diabetes
    FROM PastHistory
        WHERE ItemCode IN (3, 775, 776, 778, 774, 7840, 11998)
        AND RecordStatus = 1)

AND InternalID IN (SELECT InternalID
    FROM Invoices WHERE InvoiceID IN (SELECT InvoiceID
        FROM Services
            WHERE Recordstatus = 1
            AND Servicedate > DateAdd(Month, -6, '%s'))) 

AND InternalID IN (SELECT InternalID
    FROM Invoices WHERE InvoiceID IN (SELECT InvoiceID 
        FROM Services WHERE Recordstatus = 1 
            AND Mbsitem IN (2517, 2521, 2525) 
            AND Servicedate BETWEEN DateAdd(Year, -1, '%s') AND '%s'))

ORDER BY surname, firstname
"

for (mydate in c("20190601", "20190701")) {
  cat(sprintf(sql.query, mydate, mydate, mydate, mydate, mydate))
  cat("\n")
}

Identify booked patients without a record (billing) of completed cycle of care

SELECT *
FROM BPS_Patients

WHERE StatusText = 'Active'

AND InternalID IN (SELECT InternalID 
    FROM Appointments
    INNER JOIN (VALUES('%bhagwat%'),('%fong%'),('%ekanayake%'),('%shoesmith%'), 
            ('%plastow%'),('%samarawickrama%'),('%obeyesekere%'),('%chaves%'),
            ('%ryan%'),('%mikhail%'),('%haynes%'),('%buckwell%'),('%maxwell%'),
            ('%grace ho%'),('%bullen%'),('%lambrou%'),('%zeigler%')) AS providers(Name)
    ON dbo.BPSPayee(UserID) LIKE providers.Name

    WHERE AppointmentDate = '20190318'
    AND RecordStatus = 1
    )
    
AND InternalID IN (SELECT InternalID
    -- history of diabetes
    FROM PastHistory
        WHERE ItemCode IN (3, 775, 776, 778, 774, 7840, 11998)
        AND RecordStatus = 1)

AND InternalID NOT IN (SELECT InternalID
    -- DiabetesSIP claimed in past year
    FROM Invoices WHERE InvoiceID IN (SELECT InvoiceID 
        FROM Services WHERE Recordstatus = 1 
            AND Mbsitem IN (2517, 2521, 2525) 
            AND Servicedate BETWEEN DATEADD(Year, -1, GetDate()) AND GetDate()))
            
ORDER BY surname, firstname