Space situational awareness (SSA), for all its challenges, is relatively mature in LEO and GEO. In comparison, the situation beyond GEO is chaotic. No organization is charged with maintaining SSA for deep space objects either in distant Earth orbit or beyond Earth orbit. There is no formal interface between the astronomers who accidentally detect deep space objects while searching for asteroids and the astronautics community. Organizations such as JPL keep track of their own active probes but not of their discarded rocket stages nor the probes of other nations. This situation has been tenable due to the low flight rate of deep space missions to date, but that is changing with the arrival of commercial lunar missions and deep space cubesats, and the increasing number of states carrying out deep space exploration. I present a historical database of about 1000 deep space objects and argue that the time has come to plan for internationally coordinated deep space traffic management.
The Satellite Catalog [1,2] maintained by the US DoD (specifically, at the time of writing, USAF 18 SPCS) attempts to catalog (current and historical) Earth orbiting objects. Associated with each catalog entry are Two Line Element Sets (TLEs) giving mean geocentric SGP4 Keplerian elements , issued at a cadence of hours to weeks depending on the object.
The catalog is intended to be complete to about 10 cm size for objects in low Earth orbit, but is less complete at high altitudes. Most observations of low orbit objects use ground-based radar. Since radar's sensitivity falls off as the fourth power of distance, it is not useful for high orbit objects for which optical telescopes are used. Space-based optical telescopes are now coming on line to supplement these methods.
In addition to the US catalog, there is a Russian operational catalog; it is not public, but is thought to be not as complete for small debris objects. On the other hand, the Russian-led ISON network appears to be very successful for geosynchronous objects. Additionally, independent hobbyists provide orbit data for US military satellites whose orbit is not made public in the US catalog. European SSA is still at an experimental stage.
Although all of these systems have their limitations and problems, in general they provide a rather good knowledge of artificial objects in space within 50,000 km of the Earth.
In contrast to the comparatively healthy situation in near-Earth space, beyond 50,000 km no-one is responsible for keeping track of space activities. The US system does what can only be characterized as a half-hearted job on objects in deep Earth orbit. See, for example, the incident in which the European Integral satellite changed its orbit substantially, and the US carried on issuing elements based on propagation of the old orbit for many months until the satellite was accidentally found by asteroid observers .
No attempt is made to provide orbital data for objects which leave Earth orbit entirely. However, some (fewer than half) of them receive nominal catalog entries. Similarly, owner states do typically notify registrations of their deep space probes to the United Nations in accordance with the Registration Convention , but only very rarely do they comply with Art. IV 1(d) which stipulates the provision of `basic orbital parameters' (usually understood to include at least periapsis, apoapsis and inclination). There is no suggestion in the relevant article that it should apply only to Earth orbit.
Near-Earth Asteroid observers often accidentally find objects in deep Earth orbit, on Earth escape trajectories, or objects in solar orbit passing near the Earth. Such objects have apparent celestial motions similar in magnitude to asteroids of interest. There is a small but unfunded effort (notably by Gareth Williams, IAU Minor Planet Center , and Bill Gray, Project Pluto.)
Active deep space probes are of course tracked by their operators. However, once the probe's mission is over there is no system in place for public archiving of the trajectories. At JPL, the HORIZONS system  developed by Jon Giorgini provides ephemerides and orbit data for a subset of active and dead probes. The included missions are largely JPL-managed probes of the 1990s and later, with some missions from other agencies for which JPL has provided support and a handful of other objects added by popular demand.
HORIZONS is the single biggest contribution to SSA for deep space but it is far from a complete solution.
As humanity and its robot avatars spread into the solar system for the first time, ensuring the existence of accurate historical records has its own values. But there are more immediate reasons why the deep space catalog is needed.
Artificial deep space objects are already causing problems for astronomers. As noted above, a subset of them can be mistaken for asteroids - indeed, several were accidentally cataloged as such before the mistake was noticed and the asteroid designation retracted. Asteroid J002E3 was found in an unusual solar orbit in 2002, and was temporarily captured by the Earth-Moon system. Observations  suggested that it was actually the upper stage of the Saturn V rocket which launched Apollo 12. Spacecraft in the Sun-Earth L2 region are especially prone from this as they lie near local midnight as seen from Earth; but true solar-orbiting spacecraft passing near the Earth have also been seen. Such objects tend to have relatively low Earth-relative velocities - what a nightmare it would be if one were accidentally selected as the target of an expensive asteroid sample return mission!
Often, the presence of non-gravitational forces such as venting of residual propellant mean that state vectors or orbital elements for artificial objects generated shortly after launch are not adequate to predict the position of the object decades after launch. Nevertheless, they may be sufficient to perform a linkage if the object is serendipitously recovered: the new observations can be propagated backwards and shown to be consistent with the original orbit. Therefore, even approximate trajectory information can be helpful in confirming or ruling out proposed identifications and so space agencies should be encouraged to provide them.
Looking slightly ahead, more and more nations are sending spacecraft beyond Earth orbit, and commercial deep space missions are already beginning. Even if asteroid mining doesn't take off, we may expect that in 20 years time the entire inner solar system will be like Earth orbit today: a busy neighborhood with both scientific and commercial activities and extensive navigation and communications infrastructure. This enviroment will need governance, and governance requires situational awareness.
There is already a limited governance framework in place beyond the Outer Space Treaty. In addition to the Registration Convention already mentioned, planetary protection recommendations  are largely honoured by civil government space missions. Commercial missions, in contrast, are raising concern in this respect [11,12].
An early catalog of deep space objects  was created by the UK's Royal Aircraft Establishment (later Defense Research Agency) in 1966 as RAE Technical Report 66103, and updated a number of times [17,18,19] most recently in 1993 as the DRA Table of Space Vehicles 1958-1991 . Another early effort was a series of tables published by G. Falworth in Spaceflight and JBIS [21-26]. The present more detailed work is indebted to those earlier studies.
Other works have listed deep space missions, but generally don't include orbital data and only list active payloads. See in particluar Siddiqi's Deep Space Chronicle .