WF-24 — Partnership and Agreement Lifecycle

This workflow's own KPI is the smallest thing it does. KPI 3.4 is worth 3.0 points and, at programme level, is a binary that IoL will most likely not clear, in which case the weight is redistributed and the honest answer costs nothing. What matters is everything else this workflow owns: the agreements that four other KPIs depend on for their evidence. KPI 4.3 and KPI 5.4 both require that the partner is aware of the collaboration, and an agreement or MoU reference is how that is evidenced. KPI 5.3 counts only mobility and joint-degree options under agreements pre-approved by the CAA, so an uncovered agreement means a real student exchange scores zero. WF-13 cannot allocate a learner to a placement site without a current site agreement. KPI 6.2 counts community events organised in partnership with community organisations, which needs a documented arrangement.

The principal failure mode on KPI 3.4 itself is double counting. One industry research grant can appear in Institute - Financials.xlsx, in Institute - Research Projects.xlsx and in Institute Partnerships.xlsx, and summing all three produces a number that is wrong in the direction that looks best. This workflow defines one authoritative source per contribution type and reconciles the others to it. That is cross-KPI assertion C9.

1. Purpose

To ensure that every relationship IoL enters into with an external organisation is classified against the OBEF approved partner types at the point of signature rather than at the point of reporting, that every contribution flowing from it is valued, evidenced and recorded once against a single authoritative source, and that the agreement instrument itself exists, is current, and is retrievable, so that the four other KPIs which depend on an agreement for their evidence are able to stand up an evidence request rather than an assertion.

2. Scope

Scope statement. This process manages external partnerships and agreements from the identification of a prospective partner, through classification, negotiation, approval, signature and registration, to contribution valuation, annual reconciliation, renewal or expiry, and the assembly of the KPI 3.4 total.

Applies to. Every formal external relationship IoL enters into or relies upon: memoranda of understanding; collaboration and consortium agreements; research and industry funding agreements; consultancy, contract research and commissioned training contracts; sponsorship and donation agreements; in-kind contribution arrangements; placement and clinical training site agreements; academic mobility, joint-degree and dual-degree agreements; professional association affiliations; community-organisation arrangements underpinning volunteering and outreach; and technology partnerships involving genuine joint development. It applies to partners inside Dubai Health as well as outside it, because an internal relationship inside an academic health system is still a relationship with a distinct legal entity and is still evidenced by an agreement [IoL to confirm which Dubai Health entities are separate signing entities and which are internal to MBRU].

Does not apply to. Procurement and vendor contracts where IoL buys goods or services on ordinary commercial terms, which are not partnerships and where the supplier is not a partner. The OBEF partner-type list is explicit on the parallel case: a technology service provider counts only where there is genuine partnership and joint development, and excludes vendors who merely supply technology. Employment contracts. The research project record itself, including the 4.3 and 5.4 qualification tests and the launch-year register, which are WF-17 and WF-15; this workflow supplies the agreement reference and the partner classification that those workflows depend on and does not compute their KPIs. The course-level co-delivery claim under KPI 3.3, which is WF-03. The event itself under KPI 6.2, which is WF-23. The CE and CME commercial relationship's delivery side, which is WF-21; this workflow classifies and values the contribution arising from it.

Applicable requirements. OBEF guide v11.5 KPI 3.4 and its Appendix A anchors, together with the partner and contribution type lists, and the agreement-dependent provisions of KPI 3.3, KPI 4.3, KPI 5.3 and KPI 5.4; Appendix B evidence provisions covering partnership documents and contribution valuations; CHEDS Institute Partnerships.xlsx, Institute - Financials.xlsx, Institute - Academic Programs.xlsx, Institute - Research Projects.xlsx, Institute - Employers.xlsx and Students - Scholarship.xlsx; HEDB lookups Partner Type and Partnership Type; CAA pre-approval requirements for dual, joint and mobility agreements; MBRU delegation of authority and contract signature policy [IoL to confirm]; MBRU finance policy on donation and in-kind valuation [IoL to confirm]; data protection obligations for partner contact data, which Appendix B makes demandable.

The definitions that govern everything below.

The eight accepted contribution types for KPI 3.4. KPI 3.4 is an absolute AED total, not a ratio: direct cash value plus in-kind contributions valued at assessed market value.

The approved partner types, which must be assigned at signature. The classification is not administrative tidiness. Two of the eleven carry explicit exclusions that change whether a contribution counts at all.

The programme-level rule, which is where the honest answer lives. At institution level KPI 3.4 is scored against anchors of AED 0, 5m, 15m, 35m and 50m. At programme level it is a binary: above AED 1,000,000 over the past three years scores 100; at or below that, the 3% weight is redistributed across the other Pillar 3 KPIs. The guide provides this redistribution explicitly so that a programme without industry contributions is not penalised for their absence.

3. Trigger, boundary and endpoint

4. SIPOC

5. Accountability

Process owner. IoL Director of Partnerships, or the role able to commit IoL to an external relationship, to require classification before signature, and to resolve a conflict between a partner's expectations and MBRU policy.

R responsible · A accountable · C consulted · I informed. One A per row.

Escalation.

6. Process steps

1. Identify the prospective partner and open a register row before anything is signed or started. [CAPTURE] Partner Name, Partner Website, Country, originating IoL contact, the intended scope. Serves 3.4, 3.3, 4.3, 5.3, 5.4. Opening the row at identification rather than at signature is what makes the duplicate check in step 3 possible at all.

2. Classify the partner against the approved partner types, at this point and not later. [CAPTURE] Partner Type from the HEDB Partner Type lookup, Partner Industry, Partnership Category, Partnership Type, Area of Speclization. Serves 3.4, 4.3, 5.4. Decision point, and two traps. A federal ministry or other public body is a government agency, never a non-profit organisation. A firm that sells IoL software or equipment on ordinary commercial terms is a vendor, not a technology service provider, unless there is genuine joint development. Both misclassifications inflate the partner count and both fail on the first question asked. A third point, in the opposite direction and easier to miss: a Dubai Health hospital or clinic is a healthcare institution, which is an approved partner type. A professional association in health professions education is an approved partner type. IoL's most natural partners are therefore qualifying partners, and the risk here is under-claiming through unfamiliarity with the list rather than over-claiming.

3. Check for an existing agreement and for a duplicate relationship with the same partner. [CAPTURE] the link to any existing Partnership_ID. Serves assertion C9. Decision point. Where a relationship with this partner already exists, the new activity is normally recorded against the existing agreement rather than as a second partnership. Two register rows for one relationship is how the same contribution reaches the total twice.

4. Determine the contribution type and name its single authoritative source. Before any value is recorded, decide which of the eight contribution types applies and which CHEDS template is authoritative for it under the source-of-truth rule in section 8. [CAPTURE] contribution type, authoritative source, and the identifiers by which the other sources will be reconciled to it. Serves 3.4 and assertion C9. This step is the control. Recording the authoritative source at the moment the contribution is first known is far cheaper than reconciling three templates twelve months later and trying to work out which AED 400,000 is which.

5. Negotiate terms, and secure the attributes the KPIs need while there is still negotiating room. Four attributes are worth asking for and are almost always available at no cost: a statement of the partner's contribution and its value; permission to name the partner in reporting to the regulator; a reference number; and, where the partner will contribute to teaching, the named individuals and their expected contact hours. [CAPTURE] Partnership Description, Partner contribution, HEI Contribution. Serves 3.4, 3.3, 4.3, 5.4.

6. Obtain CAA pre-approval where the agreement supports a dual degree, a joint degree or academic mobility. [CAPTURE] CAA approval reference and date. Serves 5.3. Decision point, and it is a hard stop. KPI 5.3 counts students only under agreements pre-approved by the CAA, and mobility must be worth not less than four weeks or a three-credit-hour equivalent and must count within the programme's credit hours. An exchange run under an agreement without CAA pre-approval scores zero however real it was. Where WF-01 proposes a mobility option, this workflow must confirm the approval route before any learner is registered, not afterwards.

7. Obtain signature under the MBRU delegation of authority and file the executed agreement. [CAPTURE] Agreement type, Start Date, End Date, signature date, agreement reference, storage location. Serves 3.4, 4.3, 5.3, 5.4, and WF-13's allocation precondition. The agreement reference is the object that KPI 4.3 and KPI 5.4 depend on. Both require the partner to be aware of the collaboration, and both are evidenced through Agreement/MOU Number in Institute - Research Projects.xlsx or Agreement type in Institute Partnerships.xlsx. Note that Agreement/MOU Number is a column of the Research Projects template rather than the Partnerships template, so the reference must be pushed to WF-17 rather than assumed to be visible from the partnership record.

8. Record the cash contribution against the authoritative source only. [CAPTURE] Value of PARTNER Contribution towards Partnership, Value of University Contribution towards Partnership, Year, currency and, for a foreign-currency contribution, the conversion rate and the date used. Serves 3.4. Note the CHEDS column name Value of PARTNER Contribution towards Partnership carries a double space inside it, which will break a naive column match. Bind on the literal header string.

9. Value and evidence any in-kind contribution before recording it. [CAPTURE] the valuation, its basis, its date and the valuer. Serves 3.4. Decision point. The valuation must be an audit report, an independent third-party valuation or a formal market valuation. Unvalued in-kind is excluded, not estimated. A seconded clinician's teaching time, donated simulation equipment and free use of a partner's facilities are all legitimate in-kind contributions and all require a documented market value before they can be counted. A number produced by IoL alone with no external basis is an estimate, and estimates presented as measurements are outside this workflow's boundary.

10. Apply the type-specific exclusions before the contribution enters the total. Serves 3.4.

    • Private-sector donations exclude scholarships and endowment-earmarked gifts, which are counted under their own types instead.
    • Government-entity scholarship sponsorship is included only where the sponsorship agreement evidences an intent to employ the students. Absent that evidence, exclude it. The evidence is not a CHEDS field and lives only in the agreement, which is why the agreement must be read rather than the finance ledger.
    • Non-government scholarships count in all cases.
    • Procurement spend, ordinary vendor supply and MBRU's own contribution to a partnership are never part of the KPI 3.4 numerator; the university's own contribution has its own CHEDS column and is not a partner contribution.

11. Notify every dependent workflow that the agreement exists, with its reference and dates. [CAPTURE] the notification and its date. Serves 3.3, 4.3, 5.3, 5.4, 6.2 and WF-13. An agreement nobody downstream knows about produces the same score as no agreement. The register publishes a current list of live agreements by partner and by purpose, and WF-13 reads it as an allocation precondition rather than asking by email.

12. Monitor expiry and initiate renewal at 90 days. [CAPTURE] renewal status. Serves WF-13's hard stop and the currency of the evidence base for 4.3, 5.3 and 5.4. Exception route. Where an agreement expires while learners are placed under it, this is a compliance and learner-safety matter before it is a scoring matter, and it escalates immediately.

13. Reconcile annually across the four CHEDS sources. Compare Institute Partnerships.xlsx, Institute - Financials.xlsx, Institute - Academic Programs.xlsx programme revenue columns and Institute - Research Projects.xlsx, plus Students - Scholarship.xlsx for the scholarship type, and confirm that every contribution appears in the KPI 3.4 total exactly once. [CAPTURE] the reconciliation working paper. Serves assertion C9. Resolve the fiscal-year against academic-year alignment explicitly: Institute - Financials.xlsx is keyed on Fiscal Year while the KPI is on an academic-year basis. Document the alignment rule once and apply it identically in every year of the three-year rolling average.

14. Determine the programme-level binary, and have the classification countersigned. [CAPTURE] the three-year AED total per programme code, the determination, and the rationale for OBF3.4M. Serves 3.4 at programme level. Where the total is at or below AED 1,000,000, the weight is redistributed across the other Pillar 3 KPIs. The determination is a classification decision with a scoring consequence, so it carries the second signature the charter requires from Quality and IQA.

7. OBEF data generated

Capture rule. Four things cannot be reconstructed later and must be recorded when the relationship is formed: the partner type classification as it was decided, the executed agreement with its dates and reference, the in-kind valuation with its independent basis, and the employment-intent evidence in a government sponsorship agreement. The first can in principle be redone later but never with the same information, because the person who negotiated it will have moved on. The second either exists or does not. The third cannot be created retrospectively without becoming an estimate. The fourth lives only in the agreement and is invisible to every finance system.

Reproducibility test. A second analyst can reproduce the KPI 3.4 total only if the source-of-truth table in section 8 exists and is applied, because otherwise the same contribution can be found in three templates and the analyst has no rule for deciding which to use. The current answer is therefore no, and the capture defect is not a missing field but a missing rule. That is what action R2 creates.

8. Max-score design

The realistic reading of KPI 3.4 for IoL, stated plainly. The programme-level binary at AED 1m over three years is probably not clearable from IoL's own activity, and the redistribution provision exists precisely for that case. That is not a failure and it costs nothing. The 3.0 points move to the other Pillar 3 KPIs, where IoL has genuine strength: KPI 3.2 at 8.0 points, KPI 3.1 at 5.0 and KPI 3.3 at 4.0. The disciplined response is to measure the three-year total honestly, once, and then to classify.

What is worth noticing before concluding that the answer is redistribution. Three of the eight contribution types map directly onto activity IoL already performs, and IoL may be closer to AED 1m than it assumes:

• Type 3, direct revenue from clients, explicitly includes professional courses and certificates delivered for industry. IoL's commissioned CPD and CME delivered for Dubai Health entities and other employers falls squarely here.
• Type 7, other industry contributions, explicitly includes CPD programmes and training courses.
• Type 6, in-kind contributions, includes professional services and training programmes at documented market value. Clinician time contributed to IoL teaching, partner facilities used at no charge, and donated simulation capacity are all in-kind contributions that IoL almost certainly receives and almost certainly does not value.

The in-kind point is the material one. In-kind is counted at documented market value and IoL's largest partner contribution is probably not cash at all: it is senior clinicians' teaching time, contributed by Dubai Health entities without an invoice. Valuing that properly, with an independent basis, is legitimate, is what the guide asks for, and may move the programme-level answer. It is also the thing most likely to be over-claimed, which is why step 9 requires an external valuation basis and excludes anything without one.

What this workflow must do to reach the target.

1. Build the partnership register and populate it with every live agreement IoL relies on, including the placement site agreements WF-13 depends on and the community arrangements WF-23 depends on. Start from what exists rather than from what should exist.
2. Publish the source-of-truth table below and apply it mechanically. This is the single control that prevents the principal failure mode.
3. Classify partner type at signature, using the approved list and its two exclusions, and record the classification with a date.
4. Value in-kind contributions with an independent basis, or exclude them. No third option.
5. Read every government sponsorship agreement for employment intent before any sponsored-student funding enters the total.
6. Push the agreement reference to WF-17 so that KPI 4.3 and KPI 5.4 can evidence partner awareness, noting that the field lives in the Research Projects template.
7. Secure CAA pre-approval before any mobility or joint-degree option carries a registered student, or KPI 5.3 scores zero on a real activity.
8. Measure the three-year AED total per programme once, properly, then classify the binary and have Quality and IQA countersign it.
9. Run a 90-day expiry watch and publish the live agreement list so no downstream workflow is blocked by an agreement nobody knew had lapsed.

The source-of-truth table, which is the operational form of assertion C9.

Read the table as an instruction, not a description. For each row, the authoritative source supplies the number that enters the KPI 3.4 total, and every other appearance of the same money is a cross-check that must agree, not a second contribution to be added.

[REDESIGN] actions.

Sequencing note. R2 costs two weeks and removes the single largest source of error in this workflow. R5 should follow it, because a census run without the source-of-truth rule will produce a number that cannot be defended and may be inflated by exactly the double counting R2 exists to prevent. R1 and R6 are the infrastructure everything else sits on and also serve WF-13 and WF-25, where the points are much larger.

9. Indirect strategy where data cannot be collected

Category 3 is the principal case here, and this is the workflow where the boundary around Category 3 is most likely to be tested.

Category 3, the activity genuinely does not exist: the programme-level AED 1m binary. The guide provides weight redistribution at KPI 3.4 explicitly so that a programme without industry contributions is not penalised for their absence. For PGDipHPE and MScHPE, the honest answer is quite likely that three years of industry contributions attributable to the programme do not exceed AED 1,000,000, in which case the 3.0% weight is redistributed across the other Pillar 3 KPIs and the programme loses nothing. Using the provision correctly is not avoidance; it is what the framework instructs. What it requires is accurate classification with a documented rationale, which is action R5 and the second signature in step 14.

The line, stated explicitly because this is where it would be crossed. Redistribution is reallocation, not free money. The weight moves to the other Pillar 3 KPIs; it does not multiply. A programme with genuine industry contributions above AED 1m scores 100 on a 3.0% KPI, which is strictly better than moving that 3.0% onto KPIs scoring below 100. Classifying an activity as absent in order to trigger redistribution is therefore both dishonest and, in almost every case, worse for the score. It is in the right-hand column of the charter's table and it is not a decision available to any role in this process. The correct sequence is: measure, then classify, then record the rationale. Never: prefer an outcome, then classify to reach it.

The same discipline applies in the opposite direction, and it is the more likely error at IoL. The temptation on a binary threshold is to reach it by loosening what counts: valuing contributed clinician time generously, treating a vendor as a technology partner, counting a government scholarship without reading the sponsorship agreement for employment intent, or counting MBRU's own contribution alongside the partner's. Each of those is a misstatement on a KPI whose evidence requirement is unusually hard, because Appendix B names audit reports, independent third-party valuations and formal market valuations specifically. A KPI 3.4 total assembled from internal estimates cannot survive the evidence request that the guide plainly anticipates for this KPI.

Category 4, the activity exists but is invisible, applies to in-kind contributions and to the agreement infrastructure. Two distinct cases.

The first is in-kind valuation. IoL receives real contributions that never appear in any ledger: consultant and specialist teaching time contributed by Dubai Health entities, use of partner clinical and simulation facilities, professional services provided free, and training delivered into IoL programmes by partner staff. These are explicitly within contribution type 6, they are counted at documented market value, and they are invisible because nobody has ever valued them. The response is R4, a proper external valuation basis, not a spreadsheet estimate. This is capture rather than tactics, and the condition on it is the same as everywhere else in the pack: it must survive being asked for.

The second is the agreement itself. IoL almost certainly operates today under working relationships that have no current signed instrument, or have one that nobody can locate, or have one that expired. Every one of those is a KPI 4.3, 5.3 or 5.4 claim that cannot be evidenced, and a WF-13 placement that should not have been allocated. The response is R1, a register built from what exists, and the honest expectation is that the first pass will find gaps rather than a tidy list.

Category 2 does not apply. KPI 3.4 is Track A and is computed from financial and agreement records rather than a survey, so there is no Appendix C question here.

Category 1 does not apply. MBRU submits this KPI itself.

Category 5, performance genuinely low, is a plausible and acceptable reading at institution level. The AED 50m top anchor is high for a specialist health-sciences university, and MBRU's institutional score on KPI 3.4 will be modest whatever it does. Reporting that accurately and opening a corrective action plan, if the institution wishes to grow industry income, is the whole response. It is not a reason to inflate a number that the Ministry can ask an auditor to verify.

Two ambiguities in the guide, flagged rather than resolved.

First, whether the programme-level AED 1m threshold is inclusive. The guide says "above AED 1,000,000" and describes redistribution for contributions "at or below AED 1m". A total of exactly AED 1,000,000 therefore reads as redistribution rather than as 100. The point is unlikely to bite in practice, but where a total lands within a few per cent of the threshold, record the reading applied in OBF3.4M rather than leaving the determination to look arbitrary.

Second, how a partner contribution is attributed to a programme. KPI 3.4 is scored at programme level against a three-year AED total, but the guide does not state how a contribution to the institution, or to a department serving two programmes, is attributed between them. Institute - Academic Programs.xlsx carries programme revenue columns, which implies a direct attribution rather than an apportionment. IoL should adopt a single attribution rule, record it in OBF3.4M, and hold it constant across the three years of the rolling average, since changing basis mid-series produces a number that cannot be defended. Where a contribution genuinely serves both programmes, attributing it to both would double count it in the institutional roll-up and should not be done without a Ministry answer.

Boundary check. This workflow must never:

• classify an activity as absent, or a total as below the threshold, in order to trigger redistribution;
• record an in-kind contribution without an audit report, an independent third-party valuation or a formal market valuation behind it;
• estimate a value and present the estimate as a measurement;
• count the same contribution in more than one place because it appears in more than one CHEDS template;
• count MBRU's own contribution to a partnership as a partner contribution;
• classify a federal ministry or other public body as a non-profit organisation;
• classify an ordinary supplier as a technology service provider where there is no joint development;
• classify an entity as a research centre where it is not devoted exclusively to R&D or where it sits under a university;
• include a government-entity scholarship without evidence in the sponsorship agreement of intent to employ the students;
• count a scholarship or an endowment-earmarked gift as a private-sector donation;
• record an agreement reference against a research project where no agreement exists, since KPI 4.3 and KPI 5.4 depend on that reference to evidence partner awareness;
• describe a partner as aware of a collaboration where only an individual employee is aware of it;
• count a student under KPI 5.3 on a mobility or joint-degree agreement that lacks CAA pre-approval;
• record a community-organisation partnership for KPI 6.2 where no documented arrangement exists;
• allow a placement to be allocated under an expired site agreement;
• change the programme attribution rule or the fiscal-to-academic-year alignment rule between years without recording the change and its effect on the three-year series.

10. Service standards

11. Records and evidence

Appendix B readiness. Today, IoL could probably produce signed agreements for its larger formal relationships, though possibly not within 15 working days and possibly not as a complete set [IoL to confirm]. It could not produce an independent valuation for any in-kind contribution, because none has been commissioned, and it could not produce a reconciliation demonstrating that a contribution is counted once. The exposure on this KPI is unusual in that Appendix B names the specific instruments required, being audit reports, independent third-party valuations and formal market valuations. Most KPIs in this pack are evidenced by records IoL already creates. This one is evidenced by instruments IoL must deliberately commission, which is why R4 has a three-month lead time and why a claim made before it lands is a claim that cannot be supported.

12. Risks and controls

13. Performance measures

The measure to watch is the second accuracy row. A rising KPI 3.4 total with an unchanged reconciliation discipline is more likely to be double counting than growth, and the workflow should be able to tell the difference before the Ministry asks.

14. Change control